Instructions to use TIGER-Lab/VLM2Vec-LoRA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use TIGER-Lab/VLM2Vec-LoRA with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="TIGER-Lab/VLM2Vec-LoRA", trust_remote_code=True) messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("TIGER-Lab/VLM2Vec-LoRA", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use TIGER-Lab/VLM2Vec-LoRA with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "TIGER-Lab/VLM2Vec-LoRA" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TIGER-Lab/VLM2Vec-LoRA", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/TIGER-Lab/VLM2Vec-LoRA
- SGLang
How to use TIGER-Lab/VLM2Vec-LoRA with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "TIGER-Lab/VLM2Vec-LoRA" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TIGER-Lab/VLM2Vec-LoRA", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "TIGER-Lab/VLM2Vec-LoRA" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TIGER-Lab/VLM2Vec-LoRA", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use TIGER-Lab/VLM2Vec-LoRA with Docker Model Runner:
docker model run hf.co/TIGER-Lab/VLM2Vec-LoRA
Download trainer_state.json from TIGER-Lab/VLM2Vec-LoRA: direct link, hf CLI and curl.
- Browser
- Download file 322 kB
-
https://huggingface.co/TIGER-Lab/VLM2Vec-LoRA/resolve/main/trainer_state.json
- Command line
-
hf download hf://TIGER-Lab/VLM2Vec-LoRA/trainer_state.json
-
curl -L -o trainer_state.json https://huggingface.co/TIGER-Lab/VLM2Vec-LoRA/resolve/main/trainer_state.json
322 kB
| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.091190108191654, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0015455950540958269, | |
| "grad_norm": 472.0, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 26.6042, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0030911901081916537, | |
| "grad_norm": 490.0, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 27.2068, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00463678516228748, | |
| "grad_norm": 480.0, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 26.9311, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0061823802163833074, | |
| "grad_norm": 472.0, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 26.4758, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0077279752704791345, | |
| "grad_norm": 504.0, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 27.7404, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00927357032457496, | |
| "grad_norm": 482.0, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 26.5912, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010819165378670788, | |
| "grad_norm": 484.0, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 27.0497, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.012364760432766615, | |
| "grad_norm": 482.0, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 26.375, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.013910355486862442, | |
| "grad_norm": 470.0, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 24.9608, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.015455950540958269, | |
| "grad_norm": 484.0, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 24.9394, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.017001545595054096, | |
| "grad_norm": 464.0, | |
| "learning_rate": 2.2e-06, | |
| "loss": 23.9378, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01854714064914992, | |
| "grad_norm": 480.0, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 23.005, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02009273570324575, | |
| "grad_norm": 448.0, | |
| "learning_rate": 2.6e-06, | |
| "loss": 21.464, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.021638330757341576, | |
| "grad_norm": 430.0, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 20.2247, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.023183925811437404, | |
| "grad_norm": 416.0, | |
| "learning_rate": 3e-06, | |
| "loss": 18.9013, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02472952086553323, | |
| "grad_norm": 426.0, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 17.6419, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02627511591962906, | |
| "grad_norm": 428.0, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 16.2712, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.027820710973724884, | |
| "grad_norm": 426.0, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 14.3173, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02936630602782071, | |
| "grad_norm": 416.0, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 13.1013, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.030911901081916538, | |
| "grad_norm": 366.0, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 11.5202, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03245749613601236, | |
| "grad_norm": 322.0, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 10.0744, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03400309119010819, | |
| "grad_norm": 282.0, | |
| "learning_rate": 4.4e-06, | |
| "loss": 9.0714, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.03554868624420402, | |
| "grad_norm": 207.0, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 8.0312, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03709428129829984, | |
| "grad_norm": 178.0, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 7.7583, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03863987635239567, | |
| "grad_norm": 190.0, | |
| "learning_rate": 5e-06, | |
| "loss": 7.3539, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0401854714064915, | |
| "grad_norm": 136.0, | |
| "learning_rate": 5.2e-06, | |
| "loss": 6.9746, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04173106646058733, | |
| "grad_norm": 121.5, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 6.638, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.04327666151468315, | |
| "grad_norm": 112.0, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 6.4899, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.04482225656877898, | |
| "grad_norm": 102.5, | |
| "learning_rate": 5.8e-06, | |
| "loss": 6.2407, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04636785162287481, | |
| "grad_norm": 87.0, | |
| "learning_rate": 6e-06, | |
| "loss": 5.9863, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04791344667697063, | |
| "grad_norm": 84.0, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 5.822, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.04945904173106646, | |
| "grad_norm": 73.5, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 5.7442, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05100463678516229, | |
| "grad_norm": 72.0, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 5.4181, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.05255023183925812, | |
| "grad_norm": 67.5, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 5.2732, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05409582689335394, | |
| "grad_norm": 62.5, | |
| "learning_rate": 7e-06, | |
| "loss": 5.0897, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05564142194744977, | |
| "grad_norm": 55.5, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 4.8823, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0571870170015456, | |
| "grad_norm": 62.5, | |
| "learning_rate": 7.4e-06, | |
| "loss": 4.7194, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05873261205564142, | |
| "grad_norm": 59.0, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 4.5549, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.06027820710973725, | |
| "grad_norm": 70.5, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 4.3491, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.061823802163833076, | |
| "grad_norm": 54.25, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 4.1664, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0633693972179289, | |
| "grad_norm": 53.5, | |
| "learning_rate": 8.2e-06, | |
| "loss": 3.9786, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.06491499227202473, | |
| "grad_norm": 52.25, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 3.718, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.06646058732612056, | |
| "grad_norm": 60.0, | |
| "learning_rate": 8.6e-06, | |
| "loss": 3.57, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.06800618238021638, | |
| "grad_norm": 44.5, | |
| "learning_rate": 8.8e-06, | |
| "loss": 3.4842, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0695517774343122, | |
| "grad_norm": 39.0, | |
| "learning_rate": 9e-06, | |
| "loss": 3.0784, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07109737248840804, | |
| "grad_norm": 86.5, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 3.1007, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.07264296754250386, | |
| "grad_norm": 74.5, | |
| "learning_rate": 9.4e-06, | |
| "loss": 2.9358, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.07418856259659969, | |
| "grad_norm": 39.0, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 2.932, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.07573415765069552, | |
| "grad_norm": 35.75, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 2.7691, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.07727975270479134, | |
| "grad_norm": 32.5, | |
| "learning_rate": 1e-05, | |
| "loss": 2.6701, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07882534775888717, | |
| "grad_norm": 28.875, | |
| "learning_rate": 1.02e-05, | |
| "loss": 2.4431, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.080370942812983, | |
| "grad_norm": 30.375, | |
| "learning_rate": 1.04e-05, | |
| "loss": 2.5914, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.08191653786707882, | |
| "grad_norm": 28.875, | |
| "learning_rate": 1.0600000000000002e-05, | |
| "loss": 2.4222, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.08346213292117466, | |
| "grad_norm": 27.375, | |
| "learning_rate": 1.0800000000000002e-05, | |
| "loss": 2.1741, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.08500772797527048, | |
| "grad_norm": 55.75, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 2.1799, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0865533230293663, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 2.0037, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.08809891808346214, | |
| "grad_norm": 23.0, | |
| "learning_rate": 1.14e-05, | |
| "loss": 2.0929, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.08964451313755796, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.16e-05, | |
| "loss": 1.9233, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.09119010819165378, | |
| "grad_norm": 22.625, | |
| "learning_rate": 1.18e-05, | |
| "loss": 1.8583, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.09273570324574962, | |
| "grad_norm": 27.0, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.9135, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09428129829984544, | |
| "grad_norm": 20.875, | |
| "learning_rate": 1.22e-05, | |
| "loss": 1.7249, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.09582689335394126, | |
| "grad_norm": 20.625, | |
| "learning_rate": 1.2400000000000002e-05, | |
| "loss": 1.6811, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0973724884080371, | |
| "grad_norm": 26.5, | |
| "learning_rate": 1.2600000000000001e-05, | |
| "loss": 1.72, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.09891808346213292, | |
| "grad_norm": 18.0, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 1.6351, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.10046367851622875, | |
| "grad_norm": 18.0, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 1.6122, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10200927357032458, | |
| "grad_norm": 17.625, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 1.5638, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1035548686244204, | |
| "grad_norm": 17.875, | |
| "learning_rate": 1.3400000000000002e-05, | |
| "loss": 1.489, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.10510046367851623, | |
| "grad_norm": 18.75, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 1.5691, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.10664605873261206, | |
| "grad_norm": 20.5, | |
| "learning_rate": 1.38e-05, | |
| "loss": 1.4363, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.10819165378670788, | |
| "grad_norm": 17.125, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.4473, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10973724884080371, | |
| "grad_norm": 17.625, | |
| "learning_rate": 1.4200000000000001e-05, | |
| "loss": 1.4345, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.11128284389489954, | |
| "grad_norm": 27.5, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 1.5177, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.11282843894899536, | |
| "grad_norm": 16.625, | |
| "learning_rate": 1.46e-05, | |
| "loss": 1.3802, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1143740340030912, | |
| "grad_norm": 19.625, | |
| "learning_rate": 1.48e-05, | |
| "loss": 1.3202, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.11591962905718702, | |
| "grad_norm": 27.25, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 1.3449, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11746522411128284, | |
| "grad_norm": 16.375, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 1.3505, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.11901081916537867, | |
| "grad_norm": 16.25, | |
| "learning_rate": 1.54e-05, | |
| "loss": 1.2425, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1205564142194745, | |
| "grad_norm": 15.25, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 1.2216, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.12210200927357033, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 1.58e-05, | |
| "loss": 1.2965, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.12364760432766615, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.3441, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.125193199381762, | |
| "grad_norm": 43.25, | |
| "learning_rate": 1.62e-05, | |
| "loss": 1.2836, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.1267387944358578, | |
| "grad_norm": 18.375, | |
| "learning_rate": 1.64e-05, | |
| "loss": 1.2163, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.12828438948995363, | |
| "grad_norm": 38.25, | |
| "learning_rate": 1.66e-05, | |
| "loss": 1.1794, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.12982998454404945, | |
| "grad_norm": 14.8125, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 1.1239, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.13137557959814528, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.7e-05, | |
| "loss": 1.0771, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.13292117465224113, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 1.72e-05, | |
| "loss": 1.0916, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.13446676970633695, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 1.7400000000000003e-05, | |
| "loss": 1.0952, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.13601236476043277, | |
| "grad_norm": 13.75, | |
| "learning_rate": 1.76e-05, | |
| "loss": 1.0659, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1375579598145286, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.7800000000000002e-05, | |
| "loss": 0.9821, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1391035548686244, | |
| "grad_norm": 14.75, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.079, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14064914992272023, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.8200000000000002e-05, | |
| "loss": 1.003, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.14219474497681608, | |
| "grad_norm": 19.875, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 1.1328, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1437403400309119, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.86e-05, | |
| "loss": 1.0903, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.14528593508500773, | |
| "grad_norm": 15.5, | |
| "learning_rate": 1.88e-05, | |
| "loss": 1.0609, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.14683153013910355, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.0328, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14837712519319937, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 0.9742, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.14992272024729522, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 1.94e-05, | |
| "loss": 1.0703, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.15146831530139104, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 1.034, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.15301391035548687, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 1.98e-05, | |
| "loss": 0.9953, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1545595054095827, | |
| "grad_norm": 13.5, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9211, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1561051004636785, | |
| "grad_norm": 13.9375, | |
| "learning_rate": 1.9989473684210526e-05, | |
| "loss": 1.0385, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.15765069551777433, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.9978947368421054e-05, | |
| "loss": 0.9443, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.15919629057187018, | |
| "grad_norm": 23.625, | |
| "learning_rate": 1.9968421052631582e-05, | |
| "loss": 0.9293, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.160741885625966, | |
| "grad_norm": 15.0, | |
| "learning_rate": 1.9957894736842107e-05, | |
| "loss": 1.0052, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.16228748068006182, | |
| "grad_norm": 18.125, | |
| "learning_rate": 1.994736842105263e-05, | |
| "loss": 0.9428, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.16383307573415765, | |
| "grad_norm": 14.1875, | |
| "learning_rate": 1.993684210526316e-05, | |
| "loss": 0.9634, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.16537867078825347, | |
| "grad_norm": 14.875, | |
| "learning_rate": 1.9926315789473688e-05, | |
| "loss": 1.0048, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.16692426584234932, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.9915789473684212e-05, | |
| "loss": 0.9729, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.16846986089644514, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.990526315789474e-05, | |
| "loss": 0.9173, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.17001545595054096, | |
| "grad_norm": 18.0, | |
| "learning_rate": 1.9894736842105265e-05, | |
| "loss": 1.0079, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17156105100463678, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.988421052631579e-05, | |
| "loss": 0.9352, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1731066460587326, | |
| "grad_norm": 14.125, | |
| "learning_rate": 1.9873684210526318e-05, | |
| "loss": 0.9911, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.17465224111282843, | |
| "grad_norm": 15.25, | |
| "learning_rate": 1.9863157894736846e-05, | |
| "loss": 0.9601, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.17619783616692428, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.985263157894737e-05, | |
| "loss": 0.8756, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.1777434312210201, | |
| "grad_norm": 16.75, | |
| "learning_rate": 1.9842105263157895e-05, | |
| "loss": 0.9709, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17928902627511592, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.9831578947368423e-05, | |
| "loss": 0.9185, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.18083462132921174, | |
| "grad_norm": 14.25, | |
| "learning_rate": 1.982105263157895e-05, | |
| "loss": 0.98, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.18238021638330756, | |
| "grad_norm": 13.5, | |
| "learning_rate": 1.9810526315789476e-05, | |
| "loss": 1.0079, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1839258114374034, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.98e-05, | |
| "loss": 0.9213, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.18547140649149924, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.9789473684210528e-05, | |
| "loss": 0.9095, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18701700154559506, | |
| "grad_norm": 13.75, | |
| "learning_rate": 1.9778947368421056e-05, | |
| "loss": 0.8693, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.18856259659969088, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 1.976842105263158e-05, | |
| "loss": 0.897, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.1901081916537867, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.9757894736842105e-05, | |
| "loss": 0.7896, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.19165378670788252, | |
| "grad_norm": 12.25, | |
| "learning_rate": 1.9747368421052633e-05, | |
| "loss": 0.8458, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.19319938176197837, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.9736842105263158e-05, | |
| "loss": 0.8362, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1947449768160742, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.9726315789473686e-05, | |
| "loss": 0.8465, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.19629057187017002, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.9715789473684214e-05, | |
| "loss": 0.7972, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.19783616692426584, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.970526315789474e-05, | |
| "loss": 0.9165, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.19938176197836166, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.9694736842105263e-05, | |
| "loss": 0.9013, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2009273570324575, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.968421052631579e-05, | |
| "loss": 0.8593, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.20247295208655333, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.967368421052632e-05, | |
| "loss": 0.8728, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.20401854714064915, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.9663157894736844e-05, | |
| "loss": 0.8588, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.20556414219474498, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.965263157894737e-05, | |
| "loss": 0.837, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2071097372488408, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.9642105263157897e-05, | |
| "loss": 0.8504, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.20865533230293662, | |
| "grad_norm": 15.5625, | |
| "learning_rate": 1.9631578947368425e-05, | |
| "loss": 0.86, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.21020092735703247, | |
| "grad_norm": 13.0, | |
| "learning_rate": 1.962105263157895e-05, | |
| "loss": 0.7569, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.2117465224111283, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.9610526315789474e-05, | |
| "loss": 0.8461, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2132921174652241, | |
| "grad_norm": 29.25, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 0.7988, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.21483771251931993, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.9589473684210527e-05, | |
| "loss": 0.8452, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.21638330757341576, | |
| "grad_norm": 14.625, | |
| "learning_rate": 1.9578947368421055e-05, | |
| "loss": 0.8358, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21792890262751158, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.956842105263158e-05, | |
| "loss": 0.7932, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.21947449768160743, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.9557894736842107e-05, | |
| "loss": 0.817, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.22102009273570325, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.9547368421052632e-05, | |
| "loss": 0.795, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.22256568778979907, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.953684210526316e-05, | |
| "loss": 0.7426, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2241112828438949, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.9526315789473688e-05, | |
| "loss": 0.76, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.22565687789799072, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.9515789473684213e-05, | |
| "loss": 0.8191, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.22720247295208656, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.9505263157894737e-05, | |
| "loss": 0.8164, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2287480680061824, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.9494736842105265e-05, | |
| "loss": 0.8299, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2302936630602782, | |
| "grad_norm": 16.125, | |
| "learning_rate": 1.9484210526315793e-05, | |
| "loss": 0.8695, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.23183925811437403, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.9473684210526318e-05, | |
| "loss": 0.8691, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23338485316846985, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.9463157894736843e-05, | |
| "loss": 0.8504, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.23493044822256567, | |
| "grad_norm": 15.5625, | |
| "learning_rate": 1.945263157894737e-05, | |
| "loss": 0.8583, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.23647604327666152, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.9442105263157895e-05, | |
| "loss": 0.9378, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.23802163833075735, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.9431578947368423e-05, | |
| "loss": 0.9021, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.23956723338485317, | |
| "grad_norm": 12.625, | |
| "learning_rate": 1.9421052631578948e-05, | |
| "loss": 0.8527, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.241112828438949, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.9410526315789476e-05, | |
| "loss": 0.8327, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2426584234930448, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.94e-05, | |
| "loss": 0.8003, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.24420401854714066, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.9389473684210525e-05, | |
| "loss": 0.6912, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.24574961360123648, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.9378947368421053e-05, | |
| "loss": 0.7362, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2472952086553323, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.936842105263158e-05, | |
| "loss": 0.7379, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.24884080370942813, | |
| "grad_norm": 25.625, | |
| "learning_rate": 1.9357894736842106e-05, | |
| "loss": 0.7803, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.250386398763524, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.9347368421052634e-05, | |
| "loss": 0.8244, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.25193199381761977, | |
| "grad_norm": 13.625, | |
| "learning_rate": 1.9336842105263162e-05, | |
| "loss": 0.773, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2534775888717156, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.9326315789473687e-05, | |
| "loss": 0.7818, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2550231839258114, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.931578947368421e-05, | |
| "loss": 0.825, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.25656877897990726, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.930526315789474e-05, | |
| "loss": 0.7203, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2581143740340031, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.9294736842105264e-05, | |
| "loss": 0.8392, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.2596599690880989, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.9284210526315792e-05, | |
| "loss": 0.8234, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.26120556414219476, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.9273684210526317e-05, | |
| "loss": 0.8638, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.26275115919629055, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.9263157894736845e-05, | |
| "loss": 0.7969, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2642967542503864, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.925263157894737e-05, | |
| "loss": 0.7549, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.26584234930448225, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.9242105263157894e-05, | |
| "loss": 0.7519, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.26738794435857804, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.9231578947368422e-05, | |
| "loss": 0.7572, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.2689335394126739, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.922105263157895e-05, | |
| "loss": 0.8011, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2704791344667697, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.9210526315789474e-05, | |
| "loss": 0.7712, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.27202472952086554, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 0.6951, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.2735703245749614, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.918947368421053e-05, | |
| "loss": 0.7982, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.2751159196290572, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.9178947368421055e-05, | |
| "loss": 0.6785, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.27666151468315303, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.916842105263158e-05, | |
| "loss": 0.7832, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2782071097372488, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 1.9157894736842108e-05, | |
| "loss": 0.7041, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2797527047913447, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.9147368421052632e-05, | |
| "loss": 0.7545, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.28129829984544047, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.913684210526316e-05, | |
| "loss": 0.7533, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2828438948995363, | |
| "grad_norm": 17.75, | |
| "learning_rate": 1.9126315789473685e-05, | |
| "loss": 0.8242, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.28438948995363217, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.9115789473684213e-05, | |
| "loss": 0.6681, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.28593508500772796, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.9105263157894738e-05, | |
| "loss": 0.6771, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2874806800618238, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.9094736842105262e-05, | |
| "loss": 0.7475, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.2890262751159196, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.908421052631579e-05, | |
| "loss": 0.7511, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.29057187017001546, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 1.907368421052632e-05, | |
| "loss": 0.7622, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.2921174652241113, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 1.9063157894736843e-05, | |
| "loss": 0.7271, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2936630602782071, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.9052631578947368e-05, | |
| "loss": 0.7669, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.29520865533230295, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.9042105263157896e-05, | |
| "loss": 0.6624, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.29675425038639874, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.9031578947368424e-05, | |
| "loss": 0.7241, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2982998454404946, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.902105263157895e-05, | |
| "loss": 0.6986, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.29984544049459044, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.9010526315789476e-05, | |
| "loss": 0.7209, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.30139103554868624, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.7389, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3029366306027821, | |
| "grad_norm": 16.125, | |
| "learning_rate": 1.898947368421053e-05, | |
| "loss": 0.7554, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3044822256568779, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.8978947368421054e-05, | |
| "loss": 0.7488, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.30602782071097373, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.8968421052631582e-05, | |
| "loss": 0.6887, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3075734157650695, | |
| "grad_norm": 34.25, | |
| "learning_rate": 1.8957894736842106e-05, | |
| "loss": 0.6583, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3091190108191654, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.894736842105263e-05, | |
| "loss": 0.6745, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3106646058732612, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 1.893684210526316e-05, | |
| "loss": 0.7117, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.312210200927357, | |
| "grad_norm": 12.25, | |
| "learning_rate": 1.8926315789473687e-05, | |
| "loss": 0.7723, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.31375579598145287, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.891578947368421e-05, | |
| "loss": 0.7528, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.31530139103554866, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 1.8905263157894736e-05, | |
| "loss": 0.7505, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.3168469860896445, | |
| "grad_norm": 60.0, | |
| "learning_rate": 1.8894736842105264e-05, | |
| "loss": 0.7367, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.31839258114374036, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.8884210526315792e-05, | |
| "loss": 0.7468, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.31993817619783615, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.8873684210526317e-05, | |
| "loss": 0.6802, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.321483771251932, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.886315789473684e-05, | |
| "loss": 0.8019, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3230293663060278, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.885263157894737e-05, | |
| "loss": 0.6416, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.32457496136012365, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.8842105263157898e-05, | |
| "loss": 0.7199, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3261205564142195, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.8831578947368422e-05, | |
| "loss": 0.7252, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3276661514683153, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.882105263157895e-05, | |
| "loss": 0.7604, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.32921174652241114, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.8810526315789475e-05, | |
| "loss": 0.694, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.33075734157650694, | |
| "grad_norm": 13.9375, | |
| "learning_rate": 1.88e-05, | |
| "loss": 0.7057, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3323029366306028, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.8789473684210528e-05, | |
| "loss": 0.7296, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.33384853168469864, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.8778947368421056e-05, | |
| "loss": 0.6646, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.33539412673879443, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 1.876842105263158e-05, | |
| "loss": 0.7514, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3369397217928903, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.8757894736842105e-05, | |
| "loss": 0.6781, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3384853168469861, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.8747368421052633e-05, | |
| "loss": 0.7358, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3400309119010819, | |
| "grad_norm": 22.5, | |
| "learning_rate": 1.873684210526316e-05, | |
| "loss": 0.7296, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3415765069551777, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.8726315789473686e-05, | |
| "loss": 0.7116, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.34312210200927357, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.871578947368421e-05, | |
| "loss": 0.6634, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3446676970633694, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.8705263157894738e-05, | |
| "loss": 0.7326, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.3462132921174652, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.8694736842105266e-05, | |
| "loss": 0.7085, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.34775888717156106, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.868421052631579e-05, | |
| "loss": 0.7484, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.34930448222565685, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.8673684210526316e-05, | |
| "loss": 0.649, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.3508500772797527, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.8663157894736844e-05, | |
| "loss": 0.71, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.35239567233384855, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.8652631578947368e-05, | |
| "loss": 0.6652, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.35394126738794435, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.8642105263157896e-05, | |
| "loss": 0.6122, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.3554868624420402, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.8631578947368424e-05, | |
| "loss": 0.6739, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.357032457496136, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.862105263157895e-05, | |
| "loss": 0.7886, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.35857805255023184, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.8610526315789473e-05, | |
| "loss": 0.7362, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3601236476043277, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 1.86e-05, | |
| "loss": 0.7711, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.3616692426584235, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.858947368421053e-05, | |
| "loss": 0.7046, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.36321483771251933, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.8578947368421054e-05, | |
| "loss": 0.6497, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.36476043276661513, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.856842105263158e-05, | |
| "loss": 0.6476, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.366306027820711, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.8557894736842107e-05, | |
| "loss": 0.7496, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3678516228748068, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.8547368421052635e-05, | |
| "loss": 0.7072, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.3693972179289026, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.853684210526316e-05, | |
| "loss": 0.7318, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.37094281298299847, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.8526315789473684e-05, | |
| "loss": 0.7002, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.37248840803709427, | |
| "grad_norm": 13.0, | |
| "learning_rate": 1.8515789473684212e-05, | |
| "loss": 0.7398, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.3740340030911901, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.8505263157894737e-05, | |
| "loss": 0.7129, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.3755795981452859, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.8494736842105265e-05, | |
| "loss": 0.6971, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.37712519319938176, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.8484210526315793e-05, | |
| "loss": 0.6709, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3786707882534776, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 1.8473684210526317e-05, | |
| "loss": 0.6225, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3802163833075734, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.8463157894736842e-05, | |
| "loss": 0.6671, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.38176197836166925, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.845263157894737e-05, | |
| "loss": 0.6965, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.38330757341576505, | |
| "grad_norm": 15.125, | |
| "learning_rate": 1.8442105263157898e-05, | |
| "loss": 0.6453, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3848531684698609, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.8431578947368423e-05, | |
| "loss": 0.7017, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.38639876352395675, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.8421052631578947e-05, | |
| "loss": 0.6827, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38794435857805254, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.8410526315789475e-05, | |
| "loss": 0.6371, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3894899536321484, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 0.681, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3910355486862442, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.8389473684210528e-05, | |
| "loss": 0.6629, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.39258114374034003, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.8378947368421053e-05, | |
| "loss": 0.6976, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3941267387944359, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.836842105263158e-05, | |
| "loss": 0.6869, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3956723338485317, | |
| "grad_norm": 20.75, | |
| "learning_rate": 1.8357894736842105e-05, | |
| "loss": 0.6835, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3972179289026275, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.8347368421052633e-05, | |
| "loss": 0.7118, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3987635239567233, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.8336842105263158e-05, | |
| "loss": 0.7664, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.40030911901081917, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.8326315789473686e-05, | |
| "loss": 0.653, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.401854714064915, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.831578947368421e-05, | |
| "loss": 0.6779, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4034003091190108, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.830526315789474e-05, | |
| "loss": 0.611, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.40494590417310666, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.8294736842105267e-05, | |
| "loss": 0.6194, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.40649149922720246, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.828421052631579e-05, | |
| "loss": 0.7023, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.4080370942812983, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.8273684210526316e-05, | |
| "loss": 0.6753, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.4095826893353941, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.8263157894736844e-05, | |
| "loss": 0.7059, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.41112828438948995, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.8252631578947372e-05, | |
| "loss": 0.688, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4126738794435858, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.8242105263157897e-05, | |
| "loss": 0.7569, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.4142194744976816, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.823157894736842e-05, | |
| "loss": 0.66, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.41576506955177744, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.822105263157895e-05, | |
| "loss": 0.5989, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.41731066460587324, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.8210526315789477e-05, | |
| "loss": 0.6503, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4188562596599691, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.8200000000000002e-05, | |
| "loss": 0.7197, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.42040185471406494, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.8189473684210527e-05, | |
| "loss": 0.6916, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.42194744976816073, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.8178947368421055e-05, | |
| "loss": 0.658, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4234930448222566, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.816842105263158e-05, | |
| "loss": 0.6581, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4250386398763524, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.8157894736842107e-05, | |
| "loss": 0.6791, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4265842349304482, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.8147368421052632e-05, | |
| "loss": 0.6634, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.4281298299845441, | |
| "grad_norm": 21.375, | |
| "learning_rate": 1.813684210526316e-05, | |
| "loss": 0.743, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.42967542503863987, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.8126315789473685e-05, | |
| "loss": 0.7256, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.4312210200927357, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.8115789473684213e-05, | |
| "loss": 0.6024, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.4327666151468315, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.810526315789474e-05, | |
| "loss": 0.7624, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.43431221020092736, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.8094736842105265e-05, | |
| "loss": 0.6995, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.43585780525502316, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.808421052631579e-05, | |
| "loss": 0.6469, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.437403400309119, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.8073684210526318e-05, | |
| "loss": 0.629, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.43894899536321486, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.8063157894736846e-05, | |
| "loss": 0.7263, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.44049459041731065, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.805263157894737e-05, | |
| "loss": 0.6615, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.4420401854714065, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.8042105263157895e-05, | |
| "loss": 0.6439, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.4435857805255023, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.8031578947368423e-05, | |
| "loss": 0.571, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.44513137557959814, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.8021052631578948e-05, | |
| "loss": 0.703, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.446676970633694, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.8010526315789476e-05, | |
| "loss": 0.6262, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.4482225656877898, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.6794, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.44976816074188564, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.798947368421053e-05, | |
| "loss": 0.6678, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.45131375579598143, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.7978947368421053e-05, | |
| "loss": 0.641, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.4528593508500773, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.7968421052631578e-05, | |
| "loss": 0.6154, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.45440494590417313, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.795789473684211e-05, | |
| "loss": 0.7398, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.4559505409582689, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.7947368421052634e-05, | |
| "loss": 0.7055, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4574961360123648, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.793684210526316e-05, | |
| "loss": 0.6077, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.45904173106646057, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.7926315789473686e-05, | |
| "loss": 0.736, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.4605873261205564, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.7915789473684214e-05, | |
| "loss": 0.6597, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.46213292117465227, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.790526315789474e-05, | |
| "loss": 0.5764, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.46367851622874806, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.7894736842105264e-05, | |
| "loss": 0.6833, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4652241112828439, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 1.7884210526315792e-05, | |
| "loss": 0.5959, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.4667697063369397, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.7873684210526316e-05, | |
| "loss": 0.6949, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.46831530139103555, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.7863157894736844e-05, | |
| "loss": 0.6948, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.46986089644513135, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.785263157894737e-05, | |
| "loss": 0.6381, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4714064914992272, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.7842105263157897e-05, | |
| "loss": 0.6298, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.47295208655332305, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.7831578947368422e-05, | |
| "loss": 0.6889, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.47449768160741884, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.7821052631578946e-05, | |
| "loss": 0.6346, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4760432766615147, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.7810526315789474e-05, | |
| "loss": 0.6656, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.4775888717156105, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.7800000000000002e-05, | |
| "loss": 0.635, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.47913446676970634, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.7789473684210527e-05, | |
| "loss": 0.6428, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4806800618238022, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.7778947368421055e-05, | |
| "loss": 0.6268, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.482225656877898, | |
| "grad_norm": 16.75, | |
| "learning_rate": 1.7768421052631583e-05, | |
| "loss": 0.7004, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.48377125193199383, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.7757894736842108e-05, | |
| "loss": 0.6226, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.4853168469860896, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.7747368421052632e-05, | |
| "loss": 0.5717, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.4868624420401855, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 1.773684210526316e-05, | |
| "loss": 0.5794, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.4884080370942813, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.7726315789473685e-05, | |
| "loss": 0.6375, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.4899536321483771, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.7715789473684213e-05, | |
| "loss": 0.6641, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.49149922720247297, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.7705263157894738e-05, | |
| "loss": 0.7167, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.49304482225656876, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.7694736842105266e-05, | |
| "loss": 0.6788, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.4945904173106646, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.768421052631579e-05, | |
| "loss": 0.6849, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.49613601236476046, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.7673684210526315e-05, | |
| "loss": 0.6135, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.49768160741885625, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.7663157894736843e-05, | |
| "loss": 0.6388, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.4992272024729521, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.765263157894737e-05, | |
| "loss": 0.5765, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.500772797527048, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.7642105263157896e-05, | |
| "loss": 0.6513, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.5023183925811437, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.763157894736842e-05, | |
| "loss": 0.7304, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5038639876352395, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.7621052631578948e-05, | |
| "loss": 0.6222, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5054095826893354, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.7610526315789476e-05, | |
| "loss": 0.6632, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.5069551777434312, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.76e-05, | |
| "loss": 0.6116, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.508500772797527, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 1.758947368421053e-05, | |
| "loss": 0.7024, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.5100463678516228, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.7578947368421054e-05, | |
| "loss": 0.599, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5115919629057187, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.756842105263158e-05, | |
| "loss": 0.6844, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.5131375579598145, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.7557894736842106e-05, | |
| "loss": 0.7101, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.5146831530139103, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.7547368421052634e-05, | |
| "loss": 0.6213, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.5162287480680062, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.753684210526316e-05, | |
| "loss": 0.565, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.517774343122102, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.7526315789473683e-05, | |
| "loss": 0.6775, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5193199381761978, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.751578947368421e-05, | |
| "loss": 0.6364, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5208655332302936, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.750526315789474e-05, | |
| "loss": 0.7518, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.5224111282843895, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.7494736842105264e-05, | |
| "loss": 0.6518, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.5239567233384853, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.748421052631579e-05, | |
| "loss": 0.6856, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.5255023183925811, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.7473684210526317e-05, | |
| "loss": 0.6262, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.527047913446677, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.7463157894736845e-05, | |
| "loss": 0.6176, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.5285935085007728, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.745263157894737e-05, | |
| "loss": 0.5422, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.5301391035548686, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.7442105263157894e-05, | |
| "loss": 0.6596, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.5316846986089645, | |
| "grad_norm": 14.5, | |
| "learning_rate": 1.7431578947368422e-05, | |
| "loss": 0.672, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5332302936630603, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.742105263157895e-05, | |
| "loss": 0.6334, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5347758887171561, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.7410526315789475e-05, | |
| "loss": 0.7048, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.5363214837712519, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.7400000000000003e-05, | |
| "loss": 0.6699, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.5378670788253478, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.7389473684210527e-05, | |
| "loss": 0.682, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5394126738794436, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.7378947368421052e-05, | |
| "loss": 0.624, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5409582689335394, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.736842105263158e-05, | |
| "loss": 0.5733, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5425038639876353, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.7357894736842108e-05, | |
| "loss": 0.6272, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.5440494590417311, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.7347368421052633e-05, | |
| "loss": 0.6308, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.5455950540958269, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.7336842105263157e-05, | |
| "loss": 0.559, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5471406491499228, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.7326315789473685e-05, | |
| "loss": 0.6935, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.5486862442040186, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.7315789473684213e-05, | |
| "loss": 0.6228, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5502318392581144, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.7305263157894738e-05, | |
| "loss": 0.5491, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5517774343122102, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.7294736842105263e-05, | |
| "loss": 0.6608, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5533230293663061, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.728421052631579e-05, | |
| "loss": 0.6535, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.5548686244204019, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.727368421052632e-05, | |
| "loss": 0.6686, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5564142194744977, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.7263157894736843e-05, | |
| "loss": 0.6258, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5579598145285936, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.725263157894737e-05, | |
| "loss": 0.6164, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.5595054095826894, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.7242105263157896e-05, | |
| "loss": 0.6558, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5610510046367851, | |
| "grad_norm": 16.375, | |
| "learning_rate": 1.723157894736842e-05, | |
| "loss": 0.6748, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5625965996908809, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.722105263157895e-05, | |
| "loss": 0.6224, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5641421947449768, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.7210526315789477e-05, | |
| "loss": 0.5964, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5656877897990726, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.72e-05, | |
| "loss": 0.5954, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5672333848531684, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.7189473684210526e-05, | |
| "loss": 0.703, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5687789799072643, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.7178947368421054e-05, | |
| "loss": 0.5393, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5703245749613601, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.7168421052631582e-05, | |
| "loss": 0.5729, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5718701700154559, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.7157894736842107e-05, | |
| "loss": 0.5776, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5734157650695518, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.714736842105263e-05, | |
| "loss": 0.6248, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5749613601236476, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.713684210526316e-05, | |
| "loss": 0.6055, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5765069551777434, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.7126315789473687e-05, | |
| "loss": 0.5462, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.5780525502318392, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.7115789473684212e-05, | |
| "loss": 0.6324, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.5795981452859351, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.7105263157894737e-05, | |
| "loss": 0.5597, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5811437403400309, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.7094736842105265e-05, | |
| "loss": 0.5971, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5826893353941267, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.708421052631579e-05, | |
| "loss": 0.592, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5842349304482226, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.7073684210526317e-05, | |
| "loss": 0.5512, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5857805255023184, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.7063157894736845e-05, | |
| "loss": 0.6129, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5873261205564142, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.705263157894737e-05, | |
| "loss": 0.6398, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.58887171561051, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.7042105263157895e-05, | |
| "loss": 0.6193, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5904173106646059, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 1.7031578947368423e-05, | |
| "loss": 0.5899, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5919629057187017, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.702105263157895e-05, | |
| "loss": 0.6047, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5935085007727975, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.7010526315789475e-05, | |
| "loss": 0.6289, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5950540958268934, | |
| "grad_norm": 14.0, | |
| "learning_rate": 1.7e-05, | |
| "loss": 0.7105, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5965996908809892, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.6989473684210528e-05, | |
| "loss": 0.6865, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.598145285935085, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.6978947368421056e-05, | |
| "loss": 0.5632, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5996908809891809, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.696842105263158e-05, | |
| "loss": 0.6353, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.6012364760432767, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.6957894736842105e-05, | |
| "loss": 0.5294, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.6027820710973725, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.6947368421052633e-05, | |
| "loss": 0.5474, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6043276661514683, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.6936842105263158e-05, | |
| "loss": 0.6746, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.6058732612055642, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.6926315789473686e-05, | |
| "loss": 0.6166, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.60741885625966, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.691578947368421e-05, | |
| "loss": 0.6486, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.6089644513137558, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.690526315789474e-05, | |
| "loss": 0.6609, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.6105100463678517, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.6894736842105263e-05, | |
| "loss": 0.63, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.6120556414219475, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.688421052631579e-05, | |
| "loss": 0.6215, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.6136012364760433, | |
| "grad_norm": 20.0, | |
| "learning_rate": 1.687368421052632e-05, | |
| "loss": 0.6329, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.615146831530139, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 1.6863157894736844e-05, | |
| "loss": 0.6106, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.616692426584235, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.685263157894737e-05, | |
| "loss": 0.6392, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.6182380216383307, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.6842105263157896e-05, | |
| "loss": 0.6487, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6197836166924265, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.6831578947368424e-05, | |
| "loss": 0.6431, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.6213292117465224, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.682105263157895e-05, | |
| "loss": 0.6677, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.6228748068006182, | |
| "grad_norm": 16.0, | |
| "learning_rate": 1.6810526315789474e-05, | |
| "loss": 0.5728, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.624420401854714, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 0.6026, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.6259659969088099, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.6789473684210526e-05, | |
| "loss": 0.631, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6275115919629057, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.6778947368421054e-05, | |
| "loss": 0.6257, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.6290571870170015, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.676842105263158e-05, | |
| "loss": 0.6249, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.6306027820710973, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.6757894736842107e-05, | |
| "loss": 0.5465, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.6321483771251932, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.6747368421052632e-05, | |
| "loss": 0.5982, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.633693972179289, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.673684210526316e-05, | |
| "loss": 0.6516, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6352395672333848, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.6726315789473684e-05, | |
| "loss": 0.6128, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.6367851622874807, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.6715789473684212e-05, | |
| "loss": 0.6512, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6383307573415765, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.6705263157894737e-05, | |
| "loss": 0.6355, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6398763523956723, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.6694736842105265e-05, | |
| "loss": 0.6905, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.6414219474497682, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.6684210526315793e-05, | |
| "loss": 0.6242, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.642967542503864, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.6673684210526318e-05, | |
| "loss": 0.588, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.6445131375579598, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.6663157894736842e-05, | |
| "loss": 0.6352, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.6460587326120556, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.665263157894737e-05, | |
| "loss": 0.5938, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6476043276661515, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.66421052631579e-05, | |
| "loss": 0.6157, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.6491499227202473, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.6631578947368423e-05, | |
| "loss": 0.5659, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6506955177743431, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.6621052631578948e-05, | |
| "loss": 0.6424, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.652241112828439, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.6610526315789476e-05, | |
| "loss": 0.5827, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.6537867078825348, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.66e-05, | |
| "loss": 0.6227, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6553323029366306, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.658947368421053e-05, | |
| "loss": 0.6027, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6568778979907264, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.6578947368421053e-05, | |
| "loss": 0.5889, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6584234930448223, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.656842105263158e-05, | |
| "loss": 0.6242, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.6599690880989181, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.6557894736842106e-05, | |
| "loss": 0.5866, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6615146831530139, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.6547368421052634e-05, | |
| "loss": 0.596, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6630602782071098, | |
| "grad_norm": 14.0, | |
| "learning_rate": 1.653684210526316e-05, | |
| "loss": 0.5661, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.6646058732612056, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.6526315789473686e-05, | |
| "loss": 0.5982, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6661514683153014, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.651578947368421e-05, | |
| "loss": 0.5538, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.6676970633693973, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.650526315789474e-05, | |
| "loss": 0.648, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6692426584234931, | |
| "grad_norm": 14.9375, | |
| "learning_rate": 1.6494736842105267e-05, | |
| "loss": 0.5278, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.6707882534775889, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.648421052631579e-05, | |
| "loss": 0.5821, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6723338485316847, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.6473684210526316e-05, | |
| "loss": 0.6284, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6738794435857806, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.6463157894736844e-05, | |
| "loss": 0.6116, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6754250386398764, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.645263157894737e-05, | |
| "loss": 0.5687, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.6769706336939721, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.6442105263157897e-05, | |
| "loss": 0.5742, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.678516228748068, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.643157894736842e-05, | |
| "loss": 0.5422, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.6800618238021638, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.642105263157895e-05, | |
| "loss": 0.6112, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6816074188562596, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.6410526315789474e-05, | |
| "loss": 0.5813, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6831530139103554, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 1.64e-05, | |
| "loss": 0.5414, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.6846986089644513, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.6389473684210527e-05, | |
| "loss": 0.6109, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6862442040185471, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.6378947368421055e-05, | |
| "loss": 0.6202, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.6877897990726429, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.636842105263158e-05, | |
| "loss": 0.5505, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6893353941267388, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.6357894736842108e-05, | |
| "loss": 0.5557, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.6908809891808346, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.6347368421052636e-05, | |
| "loss": 0.5864, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.6924265842349304, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.633684210526316e-05, | |
| "loss": 0.6437, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.6939721792890263, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.6326315789473685e-05, | |
| "loss": 0.6422, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.6955177743431221, | |
| "grad_norm": 15.5, | |
| "learning_rate": 1.6315789473684213e-05, | |
| "loss": 0.5523, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6970633693972179, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.6305263157894737e-05, | |
| "loss": 0.5723, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6986089644513137, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.6294736842105265e-05, | |
| "loss": 0.6847, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.7001545595054096, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.628421052631579e-05, | |
| "loss": 0.606, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.7017001545595054, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.6273684210526318e-05, | |
| "loss": 0.6227, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.7032457496136012, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.6263157894736843e-05, | |
| "loss": 0.5474, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.7047913446676971, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.6252631578947367e-05, | |
| "loss": 0.6253, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.7063369397217929, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.6242105263157895e-05, | |
| "loss": 0.5415, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.7078825347758887, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 1.6231578947368423e-05, | |
| "loss": 0.5887, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.7094281298299846, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.6221052631578948e-05, | |
| "loss": 0.5866, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.7109737248840804, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.6210526315789473e-05, | |
| "loss": 0.6069, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7125193199381762, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.62e-05, | |
| "loss": 0.5322, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.714064914992272, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.618947368421053e-05, | |
| "loss": 0.5771, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.7156105100463679, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.6178947368421053e-05, | |
| "loss": 0.5291, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.7171561051004637, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.616842105263158e-05, | |
| "loss": 0.5948, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.7187017001545595, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.6157894736842106e-05, | |
| "loss": 0.5395, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7202472952086554, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.6147368421052634e-05, | |
| "loss": 0.524, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.7217928902627512, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.613684210526316e-05, | |
| "loss": 0.5742, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.723338485316847, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.6126315789473687e-05, | |
| "loss": 0.5524, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.7248840803709428, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.611578947368421e-05, | |
| "loss": 0.601, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.7264296754250387, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.6105263157894736e-05, | |
| "loss": 0.528, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7279752704791345, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.6094736842105264e-05, | |
| "loss": 0.6095, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.7295208655332303, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.6084210526315792e-05, | |
| "loss": 0.5767, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.7310664605873262, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.6073684210526317e-05, | |
| "loss": 0.6444, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.732612055641422, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.606315789473684e-05, | |
| "loss": 0.5567, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.7341576506955177, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.605263157894737e-05, | |
| "loss": 0.5302, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7357032457496137, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.6042105263157897e-05, | |
| "loss": 0.6625, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7372488408037094, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.6031578947368422e-05, | |
| "loss": 0.5529, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.7387944358578052, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.6021052631578947e-05, | |
| "loss": 0.5262, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.740340030911901, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.6010526315789475e-05, | |
| "loss": 0.6064, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.7418856259659969, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.5494, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7434312210200927, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.5989473684210527e-05, | |
| "loss": 0.6263, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.7449768160741885, | |
| "grad_norm": 21.75, | |
| "learning_rate": 1.5978947368421055e-05, | |
| "loss": 0.5636, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7465224111282844, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.596842105263158e-05, | |
| "loss": 0.5776, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7480680061823802, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.5957894736842105e-05, | |
| "loss": 0.4751, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.749613601236476, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.5947368421052633e-05, | |
| "loss": 0.5703, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7511591962905718, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.593684210526316e-05, | |
| "loss": 0.5683, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.7527047913446677, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.5926315789473685e-05, | |
| "loss": 0.6533, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.7542503863987635, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.591578947368421e-05, | |
| "loss": 0.563, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7557959814528593, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.5905263157894738e-05, | |
| "loss": 0.6214, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7573415765069552, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.5894736842105266e-05, | |
| "loss": 0.5328, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.758887171561051, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.588421052631579e-05, | |
| "loss": 0.6254, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.7604327666151468, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.5873684210526315e-05, | |
| "loss": 0.6091, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7619783616692427, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.5863157894736843e-05, | |
| "loss": 0.6014, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7635239567233385, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.585263157894737e-05, | |
| "loss": 0.5707, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.7650695517774343, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.5842105263157896e-05, | |
| "loss": 0.6192, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7666151468315301, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.5831578947368424e-05, | |
| "loss": 0.666, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.768160741885626, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.582105263157895e-05, | |
| "loss": 0.6, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.7697063369397218, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.5810526315789473e-05, | |
| "loss": 0.5649, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.7712519319938176, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.58e-05, | |
| "loss": 0.5569, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.7727975270479135, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 0.5915, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7743431221020093, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.5778947368421054e-05, | |
| "loss": 0.5871, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.7758887171561051, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.576842105263158e-05, | |
| "loss": 0.566, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.7774343122102009, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.5757894736842107e-05, | |
| "loss": 0.5629, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.7789799072642968, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.5747368421052635e-05, | |
| "loss": 0.6066, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.7805255023183926, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.573684210526316e-05, | |
| "loss": 0.5439, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7820710973724884, | |
| "grad_norm": 13.875, | |
| "learning_rate": 1.5726315789473684e-05, | |
| "loss": 0.5939, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.7836166924265843, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.5715789473684212e-05, | |
| "loss": 0.568, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.7851622874806801, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.570526315789474e-05, | |
| "loss": 0.5633, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.7867078825347759, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.5694736842105264e-05, | |
| "loss": 0.4905, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7882534775888718, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.568421052631579e-05, | |
| "loss": 0.5626, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7897990726429676, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.5673684210526317e-05, | |
| "loss": 0.5532, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.7913446676970634, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.5663157894736842e-05, | |
| "loss": 0.5102, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7928902627511591, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.565263157894737e-05, | |
| "loss": 0.6168, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.794435857805255, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.5642105263157898e-05, | |
| "loss": 0.5717, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7959814528593508, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.5631578947368422e-05, | |
| "loss": 0.5673, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7975270479134466, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.5621052631578947e-05, | |
| "loss": 0.5007, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.7990726429675425, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.5610526315789475e-05, | |
| "loss": 0.603, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.8006182380216383, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 0.5386, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.8021638330757341, | |
| "grad_norm": 13.25, | |
| "learning_rate": 1.5589473684210528e-05, | |
| "loss": 0.5935, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.80370942812983, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.5578947368421052e-05, | |
| "loss": 0.5865, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8052550231839258, | |
| "grad_norm": 18.375, | |
| "learning_rate": 1.556842105263158e-05, | |
| "loss": 0.6485, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.8068006182380216, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.555789473684211e-05, | |
| "loss": 0.6273, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.8083462132921174, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.5547368421052633e-05, | |
| "loss": 0.5633, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.8098918083462133, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.5536842105263158e-05, | |
| "loss": 0.5802, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.8114374034003091, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.5526315789473686e-05, | |
| "loss": 0.5109, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.8129829984544049, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.551578947368421e-05, | |
| "loss": 0.5406, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.8145285935085008, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.550526315789474e-05, | |
| "loss": 0.5531, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.8160741885625966, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.5494736842105263e-05, | |
| "loss": 0.6518, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.8176197836166924, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.548421052631579e-05, | |
| "loss": 0.5383, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.8191653786707882, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.5473684210526316e-05, | |
| "loss": 0.5739, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8207109737248841, | |
| "grad_norm": 13.875, | |
| "learning_rate": 1.5463157894736844e-05, | |
| "loss": 0.5225, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.8222565687789799, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.545263157894737e-05, | |
| "loss": 0.5946, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.8238021638330757, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.5442105263157896e-05, | |
| "loss": 0.5962, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.8253477588871716, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.543157894736842e-05, | |
| "loss": 0.534, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.8268933539412674, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.542105263157895e-05, | |
| "loss": 0.5131, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8284389489953632, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.5410526315789477e-05, | |
| "loss": 0.5951, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.8299845440494591, | |
| "grad_norm": 14.375, | |
| "learning_rate": 1.54e-05, | |
| "loss": 0.5676, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.8315301391035549, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.5389473684210526e-05, | |
| "loss": 0.5749, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.8330757341576507, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.5378947368421054e-05, | |
| "loss": 0.5402, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.8346213292117465, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.536842105263158e-05, | |
| "loss": 0.5603, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8361669242658424, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.5357894736842107e-05, | |
| "loss": 0.5902, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.8377125193199382, | |
| "grad_norm": 26.0, | |
| "learning_rate": 1.534736842105263e-05, | |
| "loss": 0.5891, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.839258114374034, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.533684210526316e-05, | |
| "loss": 0.5634, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.8408037094281299, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.5326315789473684e-05, | |
| "loss": 0.5918, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8423493044822257, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.5315789473684212e-05, | |
| "loss": 0.5903, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8438948995363215, | |
| "grad_norm": 14.1875, | |
| "learning_rate": 1.530526315789474e-05, | |
| "loss": 0.5949, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8454404945904173, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.5294736842105265e-05, | |
| "loss": 0.5468, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.8469860896445132, | |
| "grad_norm": 62.5, | |
| "learning_rate": 1.528421052631579e-05, | |
| "loss": 0.5462, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.848531684698609, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 1.5273684210526318e-05, | |
| "loss": 0.5228, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.8500772797527048, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.5263157894736846e-05, | |
| "loss": 0.592, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8516228748068007, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.525263157894737e-05, | |
| "loss": 0.5414, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8531684698608965, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.5242105263157897e-05, | |
| "loss": 0.5329, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8547140649149922, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.5231578947368421e-05, | |
| "loss": 0.5918, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8562596599690881, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.5221052631578948e-05, | |
| "loss": 0.5064, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.8578052550231839, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.5210526315789476e-05, | |
| "loss": 0.5486, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.8593508500772797, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 0.5412, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.8608964451313755, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.5189473684210526e-05, | |
| "loss": 0.5417, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.8624420401854714, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.5178947368421053e-05, | |
| "loss": 0.6457, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.8639876352395672, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.516842105263158e-05, | |
| "loss": 0.5393, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.865533230293663, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.5157894736842107e-05, | |
| "loss": 0.6142, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8670788253477589, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.5147368421052633e-05, | |
| "loss": 0.5455, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.8686244204018547, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.5136842105263158e-05, | |
| "loss": 0.5162, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.8701700154559505, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.5126315789473684e-05, | |
| "loss": 0.6003, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.8717156105100463, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.5115789473684212e-05, | |
| "loss": 0.5808, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.8732612055641422, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.5105263157894739e-05, | |
| "loss": 0.6351, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.874806800618238, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.5094736842105263e-05, | |
| "loss": 0.5996, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.8763523956723338, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.508421052631579e-05, | |
| "loss": 0.5089, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.8778979907264297, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.5073684210526316e-05, | |
| "loss": 0.4852, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.8794435857805255, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.5063157894736844e-05, | |
| "loss": 0.5551, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.8809891808346213, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.505263157894737e-05, | |
| "loss": 0.5481, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8825347758887172, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.5042105263157895e-05, | |
| "loss": 0.6355, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.884080370942813, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.5031578947368421e-05, | |
| "loss": 0.5246, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.8856259659969088, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.502105263157895e-05, | |
| "loss": 0.5812, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.8871715610510046, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.5010526315789476e-05, | |
| "loss": 0.58, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.8887171561051005, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.5265, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8902627511591963, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.4989473684210527e-05, | |
| "loss": 0.5676, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.8918083462132921, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.4978947368421053e-05, | |
| "loss": 0.5765, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.893353941267388, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.4968421052631581e-05, | |
| "loss": 0.5414, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.8948995363214838, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.4957894736842107e-05, | |
| "loss": 0.5369, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.8964451313755796, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.4947368421052632e-05, | |
| "loss": 0.5327, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8979907264296755, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.4936842105263158e-05, | |
| "loss": 0.5638, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.8995363214837713, | |
| "grad_norm": 15.625, | |
| "learning_rate": 1.4926315789473686e-05, | |
| "loss": 0.5382, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.9010819165378671, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.4915789473684213e-05, | |
| "loss": 0.551, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.9026275115919629, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.4905263157894739e-05, | |
| "loss": 0.5302, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.9041731066460588, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.4894736842105264e-05, | |
| "loss": 0.5138, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.9057187017001546, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.488421052631579e-05, | |
| "loss": 0.5535, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.9072642967542504, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.4873684210526318e-05, | |
| "loss": 0.5963, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.9088098918083463, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.4863157894736844e-05, | |
| "loss": 0.577, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.910355486862442, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.4852631578947369e-05, | |
| "loss": 0.5407, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.9119010819165378, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.4842105263157895e-05, | |
| "loss": 0.4994, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9134466769706336, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.4831578947368422e-05, | |
| "loss": 0.5298, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.9149922720247295, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.482105263157895e-05, | |
| "loss": 0.5297, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.9165378670788253, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.4810526315789476e-05, | |
| "loss": 0.5725, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.9180834621329211, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.48e-05, | |
| "loss": 0.4832, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.919629057187017, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.4789473684210527e-05, | |
| "loss": 0.5872, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.9211746522411128, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.4778947368421055e-05, | |
| "loss": 0.5893, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.9227202472952086, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 1.4768421052631581e-05, | |
| "loss": 0.5816, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.9242658423493045, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.4757894736842106e-05, | |
| "loss": 0.5638, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.9258114374034003, | |
| "grad_norm": 14.5, | |
| "learning_rate": 1.4747368421052632e-05, | |
| "loss": 0.5258, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.9273570324574961, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.4736842105263159e-05, | |
| "loss": 0.5906, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9289026275115919, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.4726315789473687e-05, | |
| "loss": 0.4826, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.9304482225656878, | |
| "grad_norm": 14.6875, | |
| "learning_rate": 1.4715789473684213e-05, | |
| "loss": 0.5613, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.9319938176197836, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.4705263157894738e-05, | |
| "loss": 0.5907, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.9335394126738794, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.4694736842105264e-05, | |
| "loss": 0.5513, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.9350850077279753, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.468421052631579e-05, | |
| "loss": 0.5197, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9366306027820711, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.4673684210526318e-05, | |
| "loss": 0.6047, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.9381761978361669, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.4663157894736843e-05, | |
| "loss": 0.6445, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.9397217928902627, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.465263157894737e-05, | |
| "loss": 0.5412, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.9412673879443586, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.4642105263157896e-05, | |
| "loss": 0.5648, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.9428129829984544, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.4631578947368424e-05, | |
| "loss": 0.6084, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9443585780525502, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.462105263157895e-05, | |
| "loss": 0.5427, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.9459041731066461, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.4610526315789474e-05, | |
| "loss": 0.5912, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9474497681607419, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.46e-05, | |
| "loss": 0.4611, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.9489953632148377, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.4589473684210527e-05, | |
| "loss": 0.5616, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.9505409582689336, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.4578947368421055e-05, | |
| "loss": 0.6371, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9520865533230294, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.456842105263158e-05, | |
| "loss": 0.5843, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9536321483771252, | |
| "grad_norm": 13.25, | |
| "learning_rate": 1.4557894736842106e-05, | |
| "loss": 0.5961, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.955177743431221, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.4547368421052632e-05, | |
| "loss": 0.5279, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.9567233384853169, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.4536842105263159e-05, | |
| "loss": 0.6456, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.9582689335394127, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.4526315789473687e-05, | |
| "loss": 0.5538, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9598145285935085, | |
| "grad_norm": 12.25, | |
| "learning_rate": 1.4515789473684211e-05, | |
| "loss": 0.5519, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.9613601236476044, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.4505263157894738e-05, | |
| "loss": 0.5315, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.9629057187017002, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 1.4494736842105264e-05, | |
| "loss": 0.5218, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.964451313755796, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.4484210526315792e-05, | |
| "loss": 0.5299, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.9659969088098919, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 1.4473684210526317e-05, | |
| "loss": 0.5595, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9675425038639877, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.4463157894736843e-05, | |
| "loss": 0.5253, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.9690880989180835, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.445263157894737e-05, | |
| "loss": 0.5111, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.9706336939721792, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.4442105263157896e-05, | |
| "loss": 0.6317, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.9721792890262752, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.4431578947368424e-05, | |
| "loss": 0.6264, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.973724884080371, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.4421052631578948e-05, | |
| "loss": 0.6112, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9752704791344667, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.4410526315789475e-05, | |
| "loss": 0.6162, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.9768160741885626, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 0.5203, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.9783616692426584, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.4389473684210526e-05, | |
| "loss": 0.5198, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.9799072642967542, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.4378947368421054e-05, | |
| "loss": 0.5171, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.98145285935085, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.436842105263158e-05, | |
| "loss": 0.5061, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9829984544049459, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.4357894736842106e-05, | |
| "loss": 0.53, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.9845440494590417, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.4347368421052633e-05, | |
| "loss": 0.5439, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.9860896445131375, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.433684210526316e-05, | |
| "loss": 0.5757, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.9876352395672334, | |
| "grad_norm": 18.25, | |
| "learning_rate": 1.4326315789473685e-05, | |
| "loss": 0.5069, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.9891808346213292, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.4315789473684212e-05, | |
| "loss": 0.6396, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.990726429675425, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.4305263157894738e-05, | |
| "loss": 0.5631, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.9922720247295209, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.4294736842105263e-05, | |
| "loss": 0.5161, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.9938176197836167, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.4284210526315792e-05, | |
| "loss": 0.5589, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.9953632148377125, | |
| "grad_norm": 12.25, | |
| "learning_rate": 1.4273684210526317e-05, | |
| "loss": 0.5475, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.9969088098918083, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.4263157894736843e-05, | |
| "loss": 0.5181, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9984544049459042, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.425263157894737e-05, | |
| "loss": 0.4886, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.4242105263157894e-05, | |
| "loss": 0.5598, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.001545595054096, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.4231578947368422e-05, | |
| "loss": 0.4375, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.0030911901081916, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.4221052631578949e-05, | |
| "loss": 0.4773, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.0046367851622875, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.4210526315789475e-05, | |
| "loss": 0.5076, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0061823802163834, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.4200000000000001e-05, | |
| "loss": 0.5444, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.007727975270479, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.418947368421053e-05, | |
| "loss": 0.4753, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.009273570324575, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.4178947368421054e-05, | |
| "loss": 0.4429, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.010819165378671, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.416842105263158e-05, | |
| "loss": 0.5413, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.0123647604327666, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.4157894736842107e-05, | |
| "loss": 0.4887, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.0139103554868625, | |
| "grad_norm": 8.375, | |
| "learning_rate": 1.4147368421052631e-05, | |
| "loss": 0.4326, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.0154559505409582, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.413684210526316e-05, | |
| "loss": 0.5413, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.017001545595054, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.4126315789473686e-05, | |
| "loss": 0.53, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.01854714064915, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.4115789473684212e-05, | |
| "loss": 0.4957, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.0200927357032457, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.4105263157894738e-05, | |
| "loss": 0.4954, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0216383307573416, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.4094736842105263e-05, | |
| "loss": 0.4645, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.0231839258114375, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.4084210526315791e-05, | |
| "loss": 0.4992, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.0247295208655331, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.4073684210526317e-05, | |
| "loss": 0.5098, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.026275115919629, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.4063157894736844e-05, | |
| "loss": 0.5931, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.027820710973725, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.4052631578947368e-05, | |
| "loss": 0.5139, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.0293663060278206, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.4042105263157896e-05, | |
| "loss": 0.4894, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.0309119010819165, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.4031578947368423e-05, | |
| "loss": 0.4719, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.0324574961360125, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.4021052631578949e-05, | |
| "loss": 0.5063, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.0340030911901081, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.4010526315789475e-05, | |
| "loss": 0.4891, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.035548686244204, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.4e-05, | |
| "loss": 0.5401, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0370942812983, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.3989473684210528e-05, | |
| "loss": 0.4784, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.0386398763523956, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.3978947368421054e-05, | |
| "loss": 0.5106, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.0401854714064915, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.396842105263158e-05, | |
| "loss": 0.4898, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.0417310664605872, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.3957894736842105e-05, | |
| "loss": 0.5321, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.0432766615146831, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.3947368421052631e-05, | |
| "loss": 0.5076, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.044822256568779, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.393684210526316e-05, | |
| "loss": 0.5249, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.0463678516228747, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.3926315789473686e-05, | |
| "loss": 0.5416, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.0479134466769706, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.3915789473684212e-05, | |
| "loss": 0.5749, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.0494590417310665, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.3905263157894737e-05, | |
| "loss": 0.4648, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.0510046367851622, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.3894736842105265e-05, | |
| "loss": 0.4699, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.052550231839258, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.3884210526315791e-05, | |
| "loss": 0.4832, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.054095826893354, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.3873684210526317e-05, | |
| "loss": 0.5309, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.0556414219474497, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.3863157894736842e-05, | |
| "loss": 0.5179, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.0571870170015456, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.3852631578947368e-05, | |
| "loss": 0.5455, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.0587326120556415, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.3842105263157896e-05, | |
| "loss": 0.5686, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.0602782071097372, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.3831578947368423e-05, | |
| "loss": 0.5095, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.061823802163833, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.3821052631578949e-05, | |
| "loss": 0.4923, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.063369397217929, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.3810526315789474e-05, | |
| "loss": 0.4592, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.0649149922720247, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.38e-05, | |
| "loss": 0.456, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.0664605873261206, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.3789473684210528e-05, | |
| "loss": 0.4445, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0680061823802163, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.3778947368421054e-05, | |
| "loss": 0.4769, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.0695517774343122, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.3768421052631579e-05, | |
| "loss": 0.5493, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.071097372488408, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.3757894736842105e-05, | |
| "loss": 0.5005, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.0726429675425038, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.3747368421052633e-05, | |
| "loss": 0.5041, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.0741885625965997, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.373684210526316e-05, | |
| "loss": 0.5143, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.0757341576506956, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 1.3726315789473686e-05, | |
| "loss": 0.4953, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.0772797527047913, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.371578947368421e-05, | |
| "loss": 0.534, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.0788253477588872, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.3705263157894737e-05, | |
| "loss": 0.479, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.080370942812983, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.3694736842105265e-05, | |
| "loss": 0.5076, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.0819165378670788, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.3684210526315791e-05, | |
| "loss": 0.505, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0834621329211747, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.3673684210526316e-05, | |
| "loss": 0.4913, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.0850077279752706, | |
| "grad_norm": 12.625, | |
| "learning_rate": 1.3663157894736842e-05, | |
| "loss": 0.4428, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.0865533230293662, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.3652631578947369e-05, | |
| "loss": 0.5485, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.0880989180834622, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.3642105263157897e-05, | |
| "loss": 0.5158, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.089644513137558, | |
| "grad_norm": 30.125, | |
| "learning_rate": 1.3631578947368423e-05, | |
| "loss": 0.4655, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.0911901081916537, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.3621052631578948e-05, | |
| "loss": 0.4943, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.0927357032457496, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.3610526315789474e-05, | |
| "loss": 0.4817, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.0942812982998453, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 0.5063, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.0958268933539412, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.3589473684210528e-05, | |
| "loss": 0.4958, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.0973724884080371, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 1.3578947368421055e-05, | |
| "loss": 0.5026, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.0989180834621328, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.356842105263158e-05, | |
| "loss": 0.4928, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.1004636785162287, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.3557894736842106e-05, | |
| "loss": 0.564, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.1020092735703246, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.3547368421052634e-05, | |
| "loss": 0.4681, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.1035548686244203, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.353684210526316e-05, | |
| "loss": 0.5466, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.1051004636785162, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.3526315789473685e-05, | |
| "loss": 0.4717, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.1066460587326121, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.3515789473684211e-05, | |
| "loss": 0.4312, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.1081916537867078, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.3505263157894737e-05, | |
| "loss": 0.5215, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.1097372488408037, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.3494736842105265e-05, | |
| "loss": 0.4868, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.1112828438948996, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.3484210526315792e-05, | |
| "loss": 0.4764, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.1128284389489953, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.3473684210526316e-05, | |
| "loss": 0.5441, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1143740340030912, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.3463157894736842e-05, | |
| "loss": 0.4406, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.1159196290571871, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.345263157894737e-05, | |
| "loss": 0.5739, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.1174652241112828, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.3442105263157897e-05, | |
| "loss": 0.5379, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.1190108191653787, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.3431578947368421e-05, | |
| "loss": 0.4805, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.1205564142194744, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.3421052631578948e-05, | |
| "loss": 0.5088, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.1221020092735703, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.3410526315789474e-05, | |
| "loss": 0.4453, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.1236476043276662, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.3400000000000002e-05, | |
| "loss": 0.5294, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.125193199381762, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.3389473684210528e-05, | |
| "loss": 0.4508, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.1267387944358578, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.3378947368421053e-05, | |
| "loss": 0.4841, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.1282843894899537, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.336842105263158e-05, | |
| "loss": 0.4764, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1298299845440494, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.3357894736842106e-05, | |
| "loss": 0.4555, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.1313755795981453, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.3347368421052634e-05, | |
| "loss": 0.5199, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.1329211746522412, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.3336842105263158e-05, | |
| "loss": 0.4891, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.1344667697063369, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.3326315789473685e-05, | |
| "loss": 0.4165, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.1360123647604328, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.3315789473684211e-05, | |
| "loss": 0.4111, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.1375579598145287, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.3305263157894739e-05, | |
| "loss": 0.5086, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.1391035548686244, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.3294736842105265e-05, | |
| "loss": 0.5237, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.1406491499227203, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.328421052631579e-05, | |
| "loss": 0.5114, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.1421947449768162, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.3273684210526316e-05, | |
| "loss": 0.4488, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.1437403400309119, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.3263157894736843e-05, | |
| "loss": 0.5056, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1452859350850078, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.325263157894737e-05, | |
| "loss": 0.4251, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.1468315301391034, | |
| "grad_norm": 20.625, | |
| "learning_rate": 1.3242105263157895e-05, | |
| "loss": 0.4894, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.1483771251931993, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.3231578947368422e-05, | |
| "loss": 0.5005, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.1499227202472952, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.3221052631578948e-05, | |
| "loss": 0.5261, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.1514683153013912, | |
| "grad_norm": 16.375, | |
| "learning_rate": 1.3210526315789476e-05, | |
| "loss": 0.5253, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.1530139103554868, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 0.5434, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.1545595054095827, | |
| "grad_norm": 21.125, | |
| "learning_rate": 1.3189473684210527e-05, | |
| "loss": 0.4824, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.1561051004636784, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.3178947368421053e-05, | |
| "loss": 0.4825, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.1576506955177743, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.316842105263158e-05, | |
| "loss": 0.5656, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.1591962905718702, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.3157894736842108e-05, | |
| "loss": 0.4975, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.160741885625966, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.3147368421052632e-05, | |
| "loss": 0.4886, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.1622874806800618, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 1.3136842105263159e-05, | |
| "loss": 0.4697, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.1638330757341577, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.3126315789473685e-05, | |
| "loss": 0.516, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.1653786707882534, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.3115789473684211e-05, | |
| "loss": 0.4741, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.1669242658423493, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.310526315789474e-05, | |
| "loss": 0.4448, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.1684698608964452, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.3094736842105264e-05, | |
| "loss": 0.4573, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.170015455950541, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.308421052631579e-05, | |
| "loss": 0.4871, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.1715610510046368, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.3073684210526317e-05, | |
| "loss": 0.4972, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.1731066460587325, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.3063157894736845e-05, | |
| "loss": 0.5072, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.1746522411128284, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.305263157894737e-05, | |
| "loss": 0.508, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1761978361669243, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.3042105263157896e-05, | |
| "loss": 0.449, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.1777434312210202, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.3031578947368422e-05, | |
| "loss": 0.4341, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.179289026275116, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.3021052631578948e-05, | |
| "loss": 0.4597, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.1808346213292118, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.3010526315789476e-05, | |
| "loss": 0.4595, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.1823802163833075, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 0.4008, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.1839258114374034, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.2989473684210527e-05, | |
| "loss": 0.4453, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.1854714064914993, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.2978947368421054e-05, | |
| "loss": 0.4873, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.187017001545595, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.2968421052631578e-05, | |
| "loss": 0.4919, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.1885625965996909, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.2957894736842108e-05, | |
| "loss": 0.4375, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.1901081916537868, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.2947368421052633e-05, | |
| "loss": 0.5009, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.1916537867078825, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.2936842105263159e-05, | |
| "loss": 0.483, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.1931993817619784, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.2926315789473685e-05, | |
| "loss": 0.4926, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.1947449768160743, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.2915789473684213e-05, | |
| "loss": 0.4601, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.19629057187017, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.2905263157894738e-05, | |
| "loss": 0.4376, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.1978361669242659, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.2894736842105264e-05, | |
| "loss": 0.4916, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.1993817619783615, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.288421052631579e-05, | |
| "loss": 0.4897, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.2009273570324575, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.2873684210526317e-05, | |
| "loss": 0.5209, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.2024729520865534, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.2863157894736845e-05, | |
| "loss": 0.5088, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.2040185471406493, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.285263157894737e-05, | |
| "loss": 0.4298, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.205564142194745, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.2842105263157896e-05, | |
| "loss": 0.4875, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2071097372488409, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.2831578947368422e-05, | |
| "loss": 0.5777, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.2086553323029365, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.2821052631578947e-05, | |
| "loss": 0.4746, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.2102009273570324, | |
| "grad_norm": 20.75, | |
| "learning_rate": 1.2810526315789475e-05, | |
| "loss": 0.4538, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.2117465224111283, | |
| "grad_norm": 13.875, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 0.4987, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.213292117465224, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.2789473684210527e-05, | |
| "loss": 0.4823, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.21483771251932, | |
| "grad_norm": 14.75, | |
| "learning_rate": 1.2778947368421054e-05, | |
| "loss": 0.5438, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.2163833075734158, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.2768421052631582e-05, | |
| "loss": 0.4995, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.2179289026275115, | |
| "grad_norm": 17.625, | |
| "learning_rate": 1.2757894736842106e-05, | |
| "loss": 0.5238, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.2194744976816074, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.2747368421052633e-05, | |
| "loss": 0.456, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.2210200927357033, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 1.2736842105263159e-05, | |
| "loss": 0.5335, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.222565687789799, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.2726315789473684e-05, | |
| "loss": 0.4543, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.224111282843895, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2715789473684212e-05, | |
| "loss": 0.5536, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.2256568778979906, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.2705263157894738e-05, | |
| "loss": 0.5724, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.2272024729520865, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.2694736842105264e-05, | |
| "loss": 0.4694, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.2287480680061824, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.268421052631579e-05, | |
| "loss": 0.5237, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.2302936630602783, | |
| "grad_norm": 16.75, | |
| "learning_rate": 1.2673684210526315e-05, | |
| "loss": 0.5096, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.231839258114374, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.2663157894736843e-05, | |
| "loss": 0.5019, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.23338485316847, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.265263157894737e-05, | |
| "loss": 0.4939, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.2349304482225656, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.2642105263157896e-05, | |
| "loss": 0.4893, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.2364760432766615, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 0.466, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2380216383307574, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.2621052631578949e-05, | |
| "loss": 0.4546, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.239567233384853, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.2610526315789475e-05, | |
| "loss": 0.4507, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.241112828438949, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.2600000000000001e-05, | |
| "loss": 0.4615, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.242658423493045, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.2589473684210528e-05, | |
| "loss": 0.5071, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.2442040185471406, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.2578947368421052e-05, | |
| "loss": 0.4829, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.2457496136012365, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.256842105263158e-05, | |
| "loss": 0.4414, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.2472952086553324, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2557894736842107e-05, | |
| "loss": 0.4959, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.248840803709428, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.2547368421052633e-05, | |
| "loss": 0.4745, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.250386398763524, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.2536842105263158e-05, | |
| "loss": 0.5139, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.2519319938176197, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.2526315789473684e-05, | |
| "loss": 0.5002, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2534775888717156, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.2515789473684212e-05, | |
| "loss": 0.478, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.2550231839258115, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.2505263157894738e-05, | |
| "loss": 0.4273, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.2565687789799074, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2494736842105265e-05, | |
| "loss": 0.4785, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.258114374034003, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.248421052631579e-05, | |
| "loss": 0.431, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.259659969088099, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.2473684210526317e-05, | |
| "loss": 0.4266, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.2612055641421946, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.2463157894736844e-05, | |
| "loss": 0.4777, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.2627511591962906, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.245263157894737e-05, | |
| "loss": 0.4552, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.2642967542503865, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.2442105263157895e-05, | |
| "loss": 0.4703, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.2658423493044824, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.2431578947368421e-05, | |
| "loss": 0.4642, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.267387944358578, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2421052631578949e-05, | |
| "loss": 0.4594, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.268933539412674, | |
| "grad_norm": 31.0, | |
| "learning_rate": 1.2410526315789475e-05, | |
| "loss": 0.5015, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.2704791344667696, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.2400000000000002e-05, | |
| "loss": 0.5548, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.2720247295208655, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.2389473684210526e-05, | |
| "loss": 0.4303, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.2735703245749614, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.2378947368421053e-05, | |
| "loss": 0.4304, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.2751159196290571, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.236842105263158e-05, | |
| "loss": 0.4988, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.276661514683153, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.2357894736842107e-05, | |
| "loss": 0.4577, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.2782071097372487, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.2347368421052631e-05, | |
| "loss": 0.422, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.2797527047913446, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.2336842105263158e-05, | |
| "loss": 0.4912, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.2812982998454405, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.2326315789473686e-05, | |
| "loss": 0.4765, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.2828438948995364, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.2315789473684212e-05, | |
| "loss": 0.5089, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2843894899536321, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.2305263157894739e-05, | |
| "loss": 0.4665, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.285935085007728, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.2294736842105263e-05, | |
| "loss": 0.4573, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.2874806800618237, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.228421052631579e-05, | |
| "loss": 0.517, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.2890262751159196, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.2273684210526317e-05, | |
| "loss": 0.4762, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.2905718701700155, | |
| "grad_norm": 13.25, | |
| "learning_rate": 1.2263157894736844e-05, | |
| "loss": 0.4777, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.2921174652241114, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.225263157894737e-05, | |
| "loss": 0.5324, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.293663060278207, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.2242105263157895e-05, | |
| "loss": 0.4963, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.295208655332303, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.2231578947368421e-05, | |
| "loss": 0.422, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.2967542503863987, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.2221052631578949e-05, | |
| "loss": 0.4485, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.2982998454404946, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.2210526315789475e-05, | |
| "loss": 0.4733, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.2998454404945905, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.22e-05, | |
| "loss": 0.4719, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.3013910355486862, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.2189473684210526e-05, | |
| "loss": 0.4764, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.302936630602782, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.2178947368421054e-05, | |
| "loss": 0.4021, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.3044822256568778, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.216842105263158e-05, | |
| "loss": 0.4517, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.3060278207109737, | |
| "grad_norm": 17.625, | |
| "learning_rate": 1.2157894736842107e-05, | |
| "loss": 0.4449, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.3075734157650696, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.2147368421052632e-05, | |
| "loss": 0.5565, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.3091190108191655, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.2136842105263158e-05, | |
| "loss": 0.4427, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.3106646058732612, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.2126315789473686e-05, | |
| "loss": 0.467, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.312210200927357, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.2115789473684212e-05, | |
| "loss": 0.5061, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.3137557959814528, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 1.2105263157894737e-05, | |
| "loss": 0.3927, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3153013910355487, | |
| "grad_norm": 14.875, | |
| "learning_rate": 1.2094736842105263e-05, | |
| "loss": 0.4699, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.3168469860896446, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.208421052631579e-05, | |
| "loss": 0.4375, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.3183925811437405, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.2073684210526318e-05, | |
| "loss": 0.4619, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.3199381761978362, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.2063157894736844e-05, | |
| "loss": 0.4803, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.321483771251932, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2052631578947369e-05, | |
| "loss": 0.5041, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.3230293663060277, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.2042105263157895e-05, | |
| "loss": 0.4666, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.3245749613601236, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.2031578947368423e-05, | |
| "loss": 0.5002, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.3261205564142196, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.202105263157895e-05, | |
| "loss": 0.5174, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.3276661514683152, | |
| "grad_norm": 14.0, | |
| "learning_rate": 1.2010526315789474e-05, | |
| "loss": 0.4862, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.3292117465224111, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.4956, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3307573415765068, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.1989473684210527e-05, | |
| "loss": 0.4812, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.3323029366306027, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.1978947368421055e-05, | |
| "loss": 0.4597, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.3338485316846986, | |
| "grad_norm": 21.25, | |
| "learning_rate": 1.1968421052631581e-05, | |
| "loss": 0.4589, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.3353941267387945, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.1957894736842106e-05, | |
| "loss": 0.4872, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.3369397217928902, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.1947368421052632e-05, | |
| "loss": 0.448, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.3384853168469861, | |
| "grad_norm": 13.0, | |
| "learning_rate": 1.1936842105263158e-05, | |
| "loss": 0.4954, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.3400309119010818, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.1926315789473686e-05, | |
| "loss": 0.5019, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.3415765069551777, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.1915789473684211e-05, | |
| "loss": 0.4909, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.3431221020092736, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.1905263157894737e-05, | |
| "loss": 0.4753, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.3446676970633695, | |
| "grad_norm": 15.5625, | |
| "learning_rate": 1.1894736842105264e-05, | |
| "loss": 0.4639, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3462132921174652, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.1884210526315792e-05, | |
| "loss": 0.459, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.3477588871715611, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.1873684210526318e-05, | |
| "loss": 0.4603, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.3493044822256568, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.1863157894736843e-05, | |
| "loss": 0.493, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.3508500772797527, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.1852631578947369e-05, | |
| "loss": 0.4714, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.3523956723338486, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.1842105263157895e-05, | |
| "loss": 0.4595, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.3539412673879443, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.1831578947368423e-05, | |
| "loss": 0.5351, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.3554868624420402, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.1821052631578948e-05, | |
| "loss": 0.5331, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.3570324574961359, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.1810526315789474e-05, | |
| "loss": 0.4049, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.3585780525502318, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.18e-05, | |
| "loss": 0.4966, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.3601236476043277, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.1789473684210527e-05, | |
| "loss": 0.5107, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3616692426584236, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.1778947368421055e-05, | |
| "loss": 0.5277, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.3632148377125193, | |
| "grad_norm": 21.75, | |
| "learning_rate": 1.176842105263158e-05, | |
| "loss": 0.454, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.3647604327666152, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.1757894736842106e-05, | |
| "loss": 0.4219, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.3663060278207109, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.1747368421052632e-05, | |
| "loss": 0.5257, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.3678516228748068, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.173684210526316e-05, | |
| "loss": 0.4527, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.3693972179289027, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.1726315789473685e-05, | |
| "loss": 0.5218, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.3709428129829986, | |
| "grad_norm": 27.125, | |
| "learning_rate": 1.1715789473684211e-05, | |
| "loss": 0.5205, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.3724884080370943, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.1705263157894737e-05, | |
| "loss": 0.535, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.3740340030911902, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.1694736842105264e-05, | |
| "loss": 0.5003, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.3755795981452859, | |
| "grad_norm": 11.875, | |
| "learning_rate": 1.1684210526315792e-05, | |
| "loss": 0.5055, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.3771251931993818, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.1673684210526316e-05, | |
| "loss": 0.4783, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.3786707882534777, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.1663157894736843e-05, | |
| "loss": 0.4727, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.3802163833075733, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 1.1652631578947369e-05, | |
| "loss": 0.4669, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.3817619783616693, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.1642105263157897e-05, | |
| "loss": 0.509, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.383307573415765, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.1631578947368423e-05, | |
| "loss": 0.5069, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.3848531684698608, | |
| "grad_norm": 12.25, | |
| "learning_rate": 1.1621052631578948e-05, | |
| "loss": 0.551, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.3863987635239567, | |
| "grad_norm": 14.875, | |
| "learning_rate": 1.1610526315789474e-05, | |
| "loss": 0.5221, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.3879443585780527, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.16e-05, | |
| "loss": 0.4923, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.3894899536321483, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.1589473684210529e-05, | |
| "loss": 0.501, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.3910355486862442, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.1578947368421053e-05, | |
| "loss": 0.4895, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.39258114374034, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.156842105263158e-05, | |
| "loss": 0.4939, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.3941267387944358, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.1557894736842106e-05, | |
| "loss": 0.4997, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.3956723338485317, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.1547368421052632e-05, | |
| "loss": 0.4545, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.3972179289026276, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.153684210526316e-05, | |
| "loss": 0.4652, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.3987635239567233, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.1526315789473685e-05, | |
| "loss": 0.4821, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.4003091190108192, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.1515789473684211e-05, | |
| "loss": 0.4238, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.401854714064915, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.1505263157894738e-05, | |
| "loss": 0.4316, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.4034003091190108, | |
| "grad_norm": 12.625, | |
| "learning_rate": 1.1494736842105266e-05, | |
| "loss": 0.5027, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.4049459041731067, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.148421052631579e-05, | |
| "loss": 0.4796, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.4064914992272024, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.1473684210526317e-05, | |
| "loss": 0.4792, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4080370942812983, | |
| "grad_norm": 8.75, | |
| "learning_rate": 1.1463157894736843e-05, | |
| "loss": 0.3923, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.409582689335394, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.145263157894737e-05, | |
| "loss": 0.516, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.41112828438949, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.1442105263157897e-05, | |
| "loss": 0.4059, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.4126738794435858, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.1431578947368422e-05, | |
| "loss": 0.509, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.4142194744976817, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.1421052631578948e-05, | |
| "loss": 0.4528, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.4157650695517774, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.1410526315789475e-05, | |
| "loss": 0.4544, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.4173106646058733, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.14e-05, | |
| "loss": 0.4991, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.418856259659969, | |
| "grad_norm": 13.0, | |
| "learning_rate": 1.1389473684210527e-05, | |
| "loss": 0.5215, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.4204018547140649, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.1378947368421054e-05, | |
| "loss": 0.4418, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.4219474497681608, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.136842105263158e-05, | |
| "loss": 0.4987, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.4234930448222567, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.1357894736842106e-05, | |
| "loss": 0.5373, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.4250386398763524, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.1347368421052634e-05, | |
| "loss": 0.4254, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.4265842349304483, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.1336842105263159e-05, | |
| "loss": 0.4403, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.428129829984544, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.1326315789473685e-05, | |
| "loss": 0.4446, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.4296754250386399, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.1315789473684212e-05, | |
| "loss": 0.4583, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.4312210200927358, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.1305263157894736e-05, | |
| "loss": 0.465, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.4327666151468315, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.1294736842105264e-05, | |
| "loss": 0.4704, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.4343122102009274, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 1.128421052631579e-05, | |
| "loss": 0.4738, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.435857805255023, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.1273684210526317e-05, | |
| "loss": 0.4362, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.437403400309119, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.1263157894736843e-05, | |
| "loss": 0.4921, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.4389489953632149, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.1252631578947368e-05, | |
| "loss": 0.4173, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.4404945904173108, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.1242105263157896e-05, | |
| "loss": 0.477, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.4420401854714064, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.1231578947368422e-05, | |
| "loss": 0.428, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.4435857805255023, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.1221052631578949e-05, | |
| "loss": 0.5187, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.445131375579598, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.1210526315789473e-05, | |
| "loss": 0.463, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.446676970633694, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 0.478, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.4482225656877898, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.1189473684210528e-05, | |
| "loss": 0.4547, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.4497681607418857, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.1178947368421054e-05, | |
| "loss": 0.5182, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.4513137557959814, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.116842105263158e-05, | |
| "loss": 0.4846, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.4528593508500773, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.1157894736842105e-05, | |
| "loss": 0.4685, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.454404945904173, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.1147368421052633e-05, | |
| "loss": 0.5032, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.455950540958269, | |
| "grad_norm": 19.125, | |
| "learning_rate": 1.1136842105263159e-05, | |
| "loss": 0.4597, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.4574961360123648, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.1126315789473685e-05, | |
| "loss": 0.5173, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.4590417310664605, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.111578947368421e-05, | |
| "loss": 0.4817, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.4605873261205564, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.1105263157894736e-05, | |
| "loss": 0.5123, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.4621329211746523, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.1094736842105264e-05, | |
| "loss": 0.4537, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.463678516228748, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.108421052631579e-05, | |
| "loss": 0.5061, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.465224111282844, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.1073684210526317e-05, | |
| "loss": 0.4714, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.4667697063369398, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.1063157894736842e-05, | |
| "loss": 0.4595, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.4683153013910355, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.105263157894737e-05, | |
| "loss": 0.485, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4698608964451314, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.1042105263157896e-05, | |
| "loss": 0.4696, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.471406491499227, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.1031578947368422e-05, | |
| "loss": 0.5382, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.472952086553323, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.1021052631578947e-05, | |
| "loss": 0.5138, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.474497681607419, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.1010526315789473e-05, | |
| "loss": 0.4927, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.4760432766615148, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 0.4716, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.4775888717156105, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.0989473684210528e-05, | |
| "loss": 0.4962, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.4791344667697064, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.0978947368421054e-05, | |
| "loss": 0.4632, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.480680061823802, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.0968421052631579e-05, | |
| "loss": 0.4632, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.482225656877898, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.0957894736842105e-05, | |
| "loss": 0.4837, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.4837712519319939, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.0947368421052633e-05, | |
| "loss": 0.4365, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4853168469860896, | |
| "grad_norm": 26.5, | |
| "learning_rate": 1.093684210526316e-05, | |
| "loss": 0.5004, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.4868624420401855, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.0926315789473686e-05, | |
| "loss": 0.4456, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.4884080370942814, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.091578947368421e-05, | |
| "loss": 0.5036, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.489953632148377, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.0905263157894738e-05, | |
| "loss": 0.4448, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.491499227202473, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.0894736842105265e-05, | |
| "loss": 0.3977, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.4930448222565689, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.0884210526315791e-05, | |
| "loss": 0.5065, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.4945904173106646, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.0873684210526316e-05, | |
| "loss": 0.4816, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.4961360123647605, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.0863157894736842e-05, | |
| "loss": 0.486, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.4976816074188561, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.085263157894737e-05, | |
| "loss": 0.4613, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.499227202472952, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 1.0842105263157896e-05, | |
| "loss": 0.4749, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.500772797527048, | |
| "grad_norm": 37.0, | |
| "learning_rate": 1.0831578947368423e-05, | |
| "loss": 0.431, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.5023183925811439, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.0821052631578947e-05, | |
| "loss": 0.4514, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.5038639876352395, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.0810526315789474e-05, | |
| "loss": 0.4485, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.5054095826893354, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.0800000000000002e-05, | |
| "loss": 0.4487, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.5069551777434311, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.0789473684210528e-05, | |
| "loss": 0.5331, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.508500772797527, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.0778947368421053e-05, | |
| "loss": 0.4387, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.510046367851623, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.0768421052631579e-05, | |
| "loss": 0.4714, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.5115919629057188, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.0757894736842107e-05, | |
| "loss": 0.4552, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.5131375579598145, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.0747368421052633e-05, | |
| "loss": 0.4952, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.5146831530139102, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.073684210526316e-05, | |
| "loss": 0.4592, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5162287480680061, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.0726315789473684e-05, | |
| "loss": 0.5091, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.517774343122102, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.071578947368421e-05, | |
| "loss": 0.4525, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.519319938176198, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.0705263157894739e-05, | |
| "loss": 0.4868, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.5208655332302936, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.0694736842105265e-05, | |
| "loss": 0.4312, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.5224111282843895, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.068421052631579e-05, | |
| "loss": 0.4687, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.5239567233384852, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.0673684210526316e-05, | |
| "loss": 0.5132, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.525502318392581, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.0663157894736842e-05, | |
| "loss": 0.4795, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.527047913446677, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.065263157894737e-05, | |
| "loss": 0.5154, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.528593508500773, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.0642105263157897e-05, | |
| "loss": 0.4829, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.5301391035548686, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.0631578947368421e-05, | |
| "loss": 0.4962, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5316846986089645, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.0621052631578948e-05, | |
| "loss": 0.529, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.5332302936630602, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.0610526315789476e-05, | |
| "loss": 0.462, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.534775888717156, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 1.0600000000000002e-05, | |
| "loss": 0.4847, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.536321483771252, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.0589473684210526e-05, | |
| "loss": 0.4255, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.537867078825348, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.0578947368421053e-05, | |
| "loss": 0.4902, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.5394126738794436, | |
| "grad_norm": 13.9375, | |
| "learning_rate": 1.0568421052631579e-05, | |
| "loss": 0.5037, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.5409582689335393, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.0557894736842107e-05, | |
| "loss": 0.5112, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.5425038639876352, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.0547368421052633e-05, | |
| "loss": 0.4961, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.544049459041731, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.0536842105263158e-05, | |
| "loss": 0.4942, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.545595054095827, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 0.4709, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5471406491499229, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.051578947368421e-05, | |
| "loss": 0.4633, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.5486862442040186, | |
| "grad_norm": 16.25, | |
| "learning_rate": 1.0505263157894739e-05, | |
| "loss": 0.4135, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.5502318392581143, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.0494736842105263e-05, | |
| "loss": 0.5058, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.5517774343122102, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.048421052631579e-05, | |
| "loss": 0.4899, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.553323029366306, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.0473684210526316e-05, | |
| "loss": 0.4655, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.554868624420402, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.0463157894736844e-05, | |
| "loss": 0.4379, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 1.5564142194744977, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.045263157894737e-05, | |
| "loss": 0.5125, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 1.5579598145285936, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.0442105263157895e-05, | |
| "loss": 0.4641, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.5595054095826892, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.0431578947368421e-05, | |
| "loss": 0.439, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 1.5610510046367851, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.0421052631578948e-05, | |
| "loss": 0.462, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.562596599690881, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.0410526315789476e-05, | |
| "loss": 0.4458, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 1.564142194744977, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.04e-05, | |
| "loss": 0.4508, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.5656877897990726, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.0389473684210527e-05, | |
| "loss": 0.4882, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.5672333848531683, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.0378947368421053e-05, | |
| "loss": 0.5, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 1.5687789799072642, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.036842105263158e-05, | |
| "loss": 0.4609, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.5703245749613601, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.0357894736842107e-05, | |
| "loss": 0.4355, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.571870170015456, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.0347368421052632e-05, | |
| "loss": 0.4909, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 1.573415765069552, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.0336842105263158e-05, | |
| "loss": 0.4217, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 1.5749613601236476, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.0326315789473685e-05, | |
| "loss": 0.4935, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 1.5765069551777433, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.0315789473684213e-05, | |
| "loss": 0.4512, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5780525502318392, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.0305263157894739e-05, | |
| "loss": 0.4376, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.5795981452859351, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.0294736842105264e-05, | |
| "loss": 0.5034, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.581143740340031, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.028421052631579e-05, | |
| "loss": 0.4651, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 1.5826893353941267, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.0273684210526316e-05, | |
| "loss": 0.4074, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.5842349304482226, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.0263157894736844e-05, | |
| "loss": 0.474, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.5857805255023183, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.0252631578947369e-05, | |
| "loss": 0.4211, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 1.5873261205564142, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.0242105263157895e-05, | |
| "loss": 0.4565, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 1.58887171561051, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.0231578947368422e-05, | |
| "loss": 0.4297, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.590417310664606, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.0221052631578948e-05, | |
| "loss": 0.5189, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.5919629057187017, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.0210526315789476e-05, | |
| "loss": 0.4161, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.5935085007727974, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.02e-05, | |
| "loss": 0.4562, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 1.5950540958268933, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.0189473684210527e-05, | |
| "loss": 0.4355, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.5965996908809892, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.0178947368421053e-05, | |
| "loss": 0.407, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 1.598145285935085, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.0168421052631581e-05, | |
| "loss": 0.4643, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 1.599690880989181, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.0157894736842106e-05, | |
| "loss": 0.4079, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.6012364760432767, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.0147368421052632e-05, | |
| "loss": 0.4584, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.6027820710973724, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.0136842105263159e-05, | |
| "loss": 0.4669, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 1.6043276661514683, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.0126315789473685e-05, | |
| "loss": 0.5169, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 1.6058732612055642, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.0115789473684213e-05, | |
| "loss": 0.4227, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.60741885625966, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.0105263157894738e-05, | |
| "loss": 0.4692, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6089644513137558, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.0094736842105264e-05, | |
| "loss": 0.4665, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 1.6105100463678517, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.008421052631579e-05, | |
| "loss": 0.465, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 1.6120556414219473, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.0073684210526315e-05, | |
| "loss": 0.4639, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.6136012364760433, | |
| "grad_norm": 10.25, | |
| "learning_rate": 1.0063157894736843e-05, | |
| "loss": 0.3995, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.6151468315301392, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.005263157894737e-05, | |
| "loss": 0.4406, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.616692426584235, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.0042105263157896e-05, | |
| "loss": 0.4526, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 1.6182380216383307, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.0031578947368422e-05, | |
| "loss": 0.428, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 1.6197836166924264, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.002105263157895e-05, | |
| "loss": 0.4434, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 1.6213292117465223, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.0010526315789474e-05, | |
| "loss": 0.4815, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 1.6228748068006182, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4534, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6244204018547141, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.989473684210527e-06, | |
| "loss": 0.432, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 1.62596599690881, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.978947368421053e-06, | |
| "loss": 0.4792, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 1.6275115919629057, | |
| "grad_norm": 14.625, | |
| "learning_rate": 9.96842105263158e-06, | |
| "loss": 0.5092, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 1.6290571870170014, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.957894736842106e-06, | |
| "loss": 0.4296, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 1.6306027820710973, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.947368421052632e-06, | |
| "loss": 0.4024, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.6321483771251932, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 9.936842105263159e-06, | |
| "loss": 0.4677, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 1.6336939721792891, | |
| "grad_norm": 10.25, | |
| "learning_rate": 9.926315789473685e-06, | |
| "loss": 0.4696, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 1.6352395672333848, | |
| "grad_norm": 10.875, | |
| "learning_rate": 9.915789473684211e-06, | |
| "loss": 0.4547, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 1.6367851622874807, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.905263157894738e-06, | |
| "loss": 0.4883, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 1.6383307573415764, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.894736842105264e-06, | |
| "loss": 0.4173, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6398763523956723, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 9.88421052631579e-06, | |
| "loss": 0.4968, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 1.6414219474497682, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 9.873684210526317e-06, | |
| "loss": 0.3705, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 1.6429675425038641, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 9.863157894736843e-06, | |
| "loss": 0.4502, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 1.6445131375579598, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 9.85263157894737e-06, | |
| "loss": 0.4769, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.6460587326120555, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.842105263157896e-06, | |
| "loss": 0.4284, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.6476043276661514, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.831578947368422e-06, | |
| "loss": 0.4869, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 1.6491499227202473, | |
| "grad_norm": 10.375, | |
| "learning_rate": 9.821052631578948e-06, | |
| "loss": 0.4387, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 1.6506955177743432, | |
| "grad_norm": 11.375, | |
| "learning_rate": 9.810526315789475e-06, | |
| "loss": 0.478, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 1.652241112828439, | |
| "grad_norm": 9.75, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.3812, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 1.6537867078825348, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 9.789473684210527e-06, | |
| "loss": 0.421, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6553323029366305, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 9.778947368421054e-06, | |
| "loss": 0.441, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 1.6568778979907264, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.76842105263158e-06, | |
| "loss": 0.4801, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 1.6584234930448223, | |
| "grad_norm": 10.875, | |
| "learning_rate": 9.757894736842106e-06, | |
| "loss": 0.4156, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 1.6599690880989182, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9.747368421052633e-06, | |
| "loss": 0.4873, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 1.6615146831530139, | |
| "grad_norm": 24.375, | |
| "learning_rate": 9.736842105263159e-06, | |
| "loss": 0.4652, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.6630602782071098, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 9.726315789473685e-06, | |
| "loss": 0.4408, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 1.6646058732612055, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.715789473684212e-06, | |
| "loss": 0.4339, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 1.6661514683153014, | |
| "grad_norm": 9.625, | |
| "learning_rate": 9.705263157894738e-06, | |
| "loss": 0.4899, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 1.6676970633693973, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.694736842105263e-06, | |
| "loss": 0.4612, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 1.6692426584234932, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.68421052631579e-06, | |
| "loss": 0.4764, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6707882534775889, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 9.673684210526317e-06, | |
| "loss": 0.4575, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 1.6723338485316845, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 9.663157894736843e-06, | |
| "loss": 0.4651, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 1.6738794435857804, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.65263157894737e-06, | |
| "loss": 0.4052, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 1.6754250386398764, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9.642105263157896e-06, | |
| "loss": 0.4619, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 1.6769706336939723, | |
| "grad_norm": 20.375, | |
| "learning_rate": 9.631578947368422e-06, | |
| "loss": 0.4103, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.6785162287480682, | |
| "grad_norm": 10.625, | |
| "learning_rate": 9.621052631578947e-06, | |
| "loss": 0.4498, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 1.6800618238021638, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 9.610526315789475e-06, | |
| "loss": 0.5129, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 1.6816074188562595, | |
| "grad_norm": 10.625, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.4292, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 1.6831530139103554, | |
| "grad_norm": 13.875, | |
| "learning_rate": 9.589473684210528e-06, | |
| "loss": 0.4277, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 1.6846986089644513, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.578947368421054e-06, | |
| "loss": 0.5244, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6862442040185472, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 9.56842105263158e-06, | |
| "loss": 0.5629, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 1.687789799072643, | |
| "grad_norm": 10.75, | |
| "learning_rate": 9.557894736842107e-06, | |
| "loss": 0.4874, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.6893353941267388, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.547368421052631e-06, | |
| "loss": 0.436, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 1.6908809891808345, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 9.53684210526316e-06, | |
| "loss": 0.4772, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 1.6924265842349304, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.526315789473684e-06, | |
| "loss": 0.4902, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.6939721792890263, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 9.515789473684212e-06, | |
| "loss": 0.4513, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.6955177743431222, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 9.505263157894738e-06, | |
| "loss": 0.5156, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 1.697063369397218, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.494736842105265e-06, | |
| "loss": 0.4841, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.6986089644513136, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9.484210526315791e-06, | |
| "loss": 0.4655, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 1.7001545595054095, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 0.4568, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7017001545595054, | |
| "grad_norm": 12.375, | |
| "learning_rate": 9.463157894736844e-06, | |
| "loss": 0.4388, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 1.7032457496136013, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9.452631578947368e-06, | |
| "loss": 0.4388, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 1.7047913446676972, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 9.442105263157896e-06, | |
| "loss": 0.4927, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 1.706336939721793, | |
| "grad_norm": 9.625, | |
| "learning_rate": 9.43157894736842e-06, | |
| "loss": 0.4393, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 1.7078825347758886, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 9.421052631578949e-06, | |
| "loss": 0.5402, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.7094281298299845, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 9.410526315789475e-06, | |
| "loss": 0.4338, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 1.7109737248840804, | |
| "grad_norm": 22.0, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.4686, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 1.7125193199381763, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.389473684210528e-06, | |
| "loss": 0.4635, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 1.714064914992272, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 9.378947368421052e-06, | |
| "loss": 0.4477, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 1.7156105100463679, | |
| "grad_norm": 14.5, | |
| "learning_rate": 9.36842105263158e-06, | |
| "loss": 0.5234, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7171561051004636, | |
| "grad_norm": 9.875, | |
| "learning_rate": 9.357894736842105e-06, | |
| "loss": 0.4074, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 1.7187017001545595, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.347368421052633e-06, | |
| "loss": 0.4459, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 1.7202472952086554, | |
| "grad_norm": 13.75, | |
| "learning_rate": 9.336842105263158e-06, | |
| "loss": 0.49, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 1.7217928902627513, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.326315789473684e-06, | |
| "loss": 0.4651, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 1.723338485316847, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 9.315789473684212e-06, | |
| "loss": 0.4525, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.7248840803709427, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.305263157894737e-06, | |
| "loss": 0.4321, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 1.7264296754250386, | |
| "grad_norm": 12.375, | |
| "learning_rate": 9.294736842105265e-06, | |
| "loss": 0.4772, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 1.7279752704791345, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.28421052631579e-06, | |
| "loss": 0.4411, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 1.7295208655332304, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 9.273684210526317e-06, | |
| "loss": 0.4489, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 1.7310664605873263, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 9.263157894736842e-06, | |
| "loss": 0.4343, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.732612055641422, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.252631578947368e-06, | |
| "loss": 0.3969, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 1.7341576506955176, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 9.242105263157896e-06, | |
| "loss": 0.4903, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 1.7357032457496135, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 9.231578947368421e-06, | |
| "loss": 0.4758, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 1.7372488408037094, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 9.221052631578949e-06, | |
| "loss": 0.3913, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 1.7387944358578054, | |
| "grad_norm": 10.5, | |
| "learning_rate": 9.210526315789474e-06, | |
| "loss": 0.4722, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.740340030911901, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 0.4681, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 1.741885625965997, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 9.189473684210526e-06, | |
| "loss": 0.3733, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 1.7434312210200926, | |
| "grad_norm": 16.0, | |
| "learning_rate": 9.178947368421053e-06, | |
| "loss": 0.4727, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 1.7449768160741885, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 9.168421052631579e-06, | |
| "loss": 0.467, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 1.7465224111282844, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.157894736842105e-06, | |
| "loss": 0.471, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7480680061823803, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9.147368421052633e-06, | |
| "loss": 0.4584, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 1.749613601236476, | |
| "grad_norm": 12.125, | |
| "learning_rate": 9.136842105263158e-06, | |
| "loss": 0.4868, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 1.7511591962905717, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.126315789473686e-06, | |
| "loss": 0.4431, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 1.7527047913446676, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 9.11578947368421e-06, | |
| "loss": 0.4722, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 1.7542503863987635, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 9.105263157894739e-06, | |
| "loss": 0.4667, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.7557959814528594, | |
| "grad_norm": 10.625, | |
| "learning_rate": 9.094736842105263e-06, | |
| "loss": 0.4947, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 1.7573415765069553, | |
| "grad_norm": 10.125, | |
| "learning_rate": 9.08421052631579e-06, | |
| "loss": 0.4612, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 1.758887171561051, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.073684210526316e-06, | |
| "loss": 0.5586, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 1.7604327666151467, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.063157894736842e-06, | |
| "loss": 0.4374, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 1.7619783616692426, | |
| "grad_norm": 10.25, | |
| "learning_rate": 9.05263157894737e-06, | |
| "loss": 0.4535, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7635239567233385, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.042105263157895e-06, | |
| "loss": 0.436, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 1.7650695517774344, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 9.031578947368423e-06, | |
| "loss": 0.5334, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 1.76661514683153, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.021052631578948e-06, | |
| "loss": 0.5028, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 1.768160741885626, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.010526315789474e-06, | |
| "loss": 0.4424, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 1.7697063369397217, | |
| "grad_norm": 9.75, | |
| "learning_rate": 9e-06, | |
| "loss": 0.4092, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.7712519319938176, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 8.989473684210527e-06, | |
| "loss": 0.4267, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 1.7727975270479135, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 8.978947368421055e-06, | |
| "loss": 0.4288, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 1.7743431221020094, | |
| "grad_norm": 11.125, | |
| "learning_rate": 8.96842105263158e-06, | |
| "loss": 0.5062, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 1.775888717156105, | |
| "grad_norm": 10.0, | |
| "learning_rate": 8.957894736842107e-06, | |
| "loss": 0.4726, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 1.7774343122102008, | |
| "grad_norm": 10.25, | |
| "learning_rate": 8.947368421052632e-06, | |
| "loss": 0.4466, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7789799072642967, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 8.936842105263158e-06, | |
| "loss": 0.5098, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 1.7805255023183926, | |
| "grad_norm": 10.25, | |
| "learning_rate": 8.926315789473685e-06, | |
| "loss": 0.4475, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 1.7820710973724885, | |
| "grad_norm": 9.75, | |
| "learning_rate": 8.915789473684211e-06, | |
| "loss": 0.425, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 1.7836166924265844, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 8.905263157894737e-06, | |
| "loss": 0.5059, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 1.78516228748068, | |
| "grad_norm": 13.75, | |
| "learning_rate": 8.894736842105264e-06, | |
| "loss": 0.4619, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.7867078825347757, | |
| "grad_norm": 10.0, | |
| "learning_rate": 8.884210526315792e-06, | |
| "loss": 0.4803, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 1.7882534775888717, | |
| "grad_norm": 11.875, | |
| "learning_rate": 8.873684210526316e-06, | |
| "loss": 0.4309, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 1.7897990726429676, | |
| "grad_norm": 10.0, | |
| "learning_rate": 8.863157894736842e-06, | |
| "loss": 0.4585, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 1.7913446676970635, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 8.852631578947369e-06, | |
| "loss": 0.4533, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 1.7928902627511591, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 8.842105263157895e-06, | |
| "loss": 0.4266, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.794435857805255, | |
| "grad_norm": 10.125, | |
| "learning_rate": 8.831578947368421e-06, | |
| "loss": 0.4381, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 1.7959814528593507, | |
| "grad_norm": 10.0, | |
| "learning_rate": 8.821052631578948e-06, | |
| "loss": 0.4596, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 1.7975270479134466, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 8.810526315789474e-06, | |
| "loss": 0.5047, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 1.7990726429675425, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.4673, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 1.8006182380216385, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 8.789473684210527e-06, | |
| "loss": 0.4945, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.8021638330757341, | |
| "grad_norm": 9.5, | |
| "learning_rate": 8.778947368421053e-06, | |
| "loss": 0.4753, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 1.80370942812983, | |
| "grad_norm": 9.375, | |
| "learning_rate": 8.76842105263158e-06, | |
| "loss": 0.4579, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 1.8052550231839257, | |
| "grad_norm": 23.75, | |
| "learning_rate": 8.757894736842106e-06, | |
| "loss": 0.4629, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 1.8068006182380216, | |
| "grad_norm": 10.25, | |
| "learning_rate": 8.747368421052632e-06, | |
| "loss": 0.4488, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 1.8083462132921175, | |
| "grad_norm": 9.5, | |
| "learning_rate": 8.736842105263158e-06, | |
| "loss": 0.4859, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8098918083462134, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 8.726315789473685e-06, | |
| "loss": 0.4414, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 1.8114374034003091, | |
| "grad_norm": 18.125, | |
| "learning_rate": 8.715789473684211e-06, | |
| "loss": 0.4149, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 1.8129829984544048, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 8.705263157894737e-06, | |
| "loss": 0.4415, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 1.8145285935085007, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 8.694736842105264e-06, | |
| "loss": 0.4448, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 1.8160741885625966, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 8.68421052631579e-06, | |
| "loss": 0.4808, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.8176197836166925, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 8.673684210526316e-06, | |
| "loss": 0.4208, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.8191653786707882, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 8.663157894736843e-06, | |
| "loss": 0.4413, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 1.820710973724884, | |
| "grad_norm": 24.125, | |
| "learning_rate": 8.652631578947369e-06, | |
| "loss": 0.4286, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 1.8222565687789798, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 8.642105263157895e-06, | |
| "loss": 0.4778, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 1.8238021638330757, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 8.631578947368422e-06, | |
| "loss": 0.3949, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8253477588871716, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 8.621052631578948e-06, | |
| "loss": 0.482, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 1.8268933539412675, | |
| "grad_norm": 9.25, | |
| "learning_rate": 8.610526315789474e-06, | |
| "loss": 0.3963, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 1.8284389489953632, | |
| "grad_norm": 12.0, | |
| "learning_rate": 8.6e-06, | |
| "loss": 0.4687, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 1.829984544049459, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 8.589473684210527e-06, | |
| "loss": 0.4124, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 1.8315301391035548, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 8.578947368421053e-06, | |
| "loss": 0.4552, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.8330757341576507, | |
| "grad_norm": 9.25, | |
| "learning_rate": 8.56842105263158e-06, | |
| "loss": 0.4079, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 1.8346213292117466, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 8.557894736842106e-06, | |
| "loss": 0.4101, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 1.8361669242658425, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 8.547368421052632e-06, | |
| "loss": 0.4624, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 1.8377125193199382, | |
| "grad_norm": 10.125, | |
| "learning_rate": 8.536842105263159e-06, | |
| "loss": 0.4779, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 1.8392581143740339, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 8.526315789473685e-06, | |
| "loss": 0.5607, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.8408037094281298, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 8.515789473684211e-06, | |
| "loss": 0.4958, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 1.8423493044822257, | |
| "grad_norm": 10.125, | |
| "learning_rate": 8.505263157894738e-06, | |
| "loss": 0.3887, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 1.8438948995363216, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 8.494736842105264e-06, | |
| "loss": 0.4486, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 1.8454404945904173, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 8.48421052631579e-06, | |
| "loss": 0.4227, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 1.8469860896445132, | |
| "grad_norm": 9.625, | |
| "learning_rate": 8.473684210526317e-06, | |
| "loss": 0.4628, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.8485316846986088, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 8.463157894736843e-06, | |
| "loss": 0.4385, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 1.8500772797527048, | |
| "grad_norm": 9.875, | |
| "learning_rate": 8.45263157894737e-06, | |
| "loss": 0.4347, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 1.8516228748068007, | |
| "grad_norm": 10.25, | |
| "learning_rate": 8.442105263157896e-06, | |
| "loss": 0.4574, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 1.8531684698608966, | |
| "grad_norm": 10.875, | |
| "learning_rate": 8.431578947368422e-06, | |
| "loss": 0.4808, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 1.8547140649149922, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 0.4708, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8562596599690881, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 8.410526315789475e-06, | |
| "loss": 0.5147, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 1.8578052550231838, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.4379, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 1.8593508500772797, | |
| "grad_norm": 11.625, | |
| "learning_rate": 8.389473684210527e-06, | |
| "loss": 0.4416, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 1.8608964451313756, | |
| "grad_norm": 10.625, | |
| "learning_rate": 8.378947368421054e-06, | |
| "loss": 0.4611, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 1.8624420401854715, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 8.36842105263158e-06, | |
| "loss": 0.4531, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.8639876352395672, | |
| "grad_norm": 10.75, | |
| "learning_rate": 8.357894736842106e-06, | |
| "loss": 0.4236, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 1.865533230293663, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 8.347368421052633e-06, | |
| "loss": 0.4655, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 1.8670788253477588, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 8.336842105263159e-06, | |
| "loss": 0.4587, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 1.8686244204018547, | |
| "grad_norm": 10.75, | |
| "learning_rate": 8.326315789473685e-06, | |
| "loss": 0.5585, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 1.8701700154559506, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 8.315789473684212e-06, | |
| "loss": 0.3895, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.8717156105100463, | |
| "grad_norm": 9.5, | |
| "learning_rate": 8.305263157894738e-06, | |
| "loss": 0.3855, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 1.8732612055641422, | |
| "grad_norm": 10.75, | |
| "learning_rate": 8.294736842105264e-06, | |
| "loss": 0.4555, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 1.874806800618238, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 8.28421052631579e-06, | |
| "loss": 0.5291, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 1.8763523956723338, | |
| "grad_norm": 10.125, | |
| "learning_rate": 8.273684210526317e-06, | |
| "loss": 0.4472, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 1.8778979907264297, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 8.263157894736843e-06, | |
| "loss": 0.463, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.8794435857805256, | |
| "grad_norm": 19.625, | |
| "learning_rate": 8.25263157894737e-06, | |
| "loss": 0.484, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 1.8809891808346213, | |
| "grad_norm": 9.875, | |
| "learning_rate": 8.242105263157896e-06, | |
| "loss": 0.4934, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 1.8825347758887172, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 8.231578947368422e-06, | |
| "loss": 0.417, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 1.8840803709428129, | |
| "grad_norm": 10.0, | |
| "learning_rate": 8.221052631578948e-06, | |
| "loss": 0.4737, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 1.8856259659969088, | |
| "grad_norm": 11.75, | |
| "learning_rate": 8.210526315789475e-06, | |
| "loss": 0.4513, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.8871715610510047, | |
| "grad_norm": 10.5, | |
| "learning_rate": 8.2e-06, | |
| "loss": 0.4611, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 1.8887171561051006, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 8.189473684210527e-06, | |
| "loss": 0.4833, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 1.8902627511591963, | |
| "grad_norm": 12.0, | |
| "learning_rate": 8.178947368421054e-06, | |
| "loss": 0.4304, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 1.891808346213292, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 8.16842105263158e-06, | |
| "loss": 0.4075, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 1.8933539412673879, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 8.157894736842106e-06, | |
| "loss": 0.5219, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.8948995363214838, | |
| "grad_norm": 8.0, | |
| "learning_rate": 8.147368421052633e-06, | |
| "loss": 0.3928, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 1.8964451313755797, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 8.136842105263159e-06, | |
| "loss": 0.398, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 1.8979907264296756, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 8.126315789473684e-06, | |
| "loss": 0.4388, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 1.8995363214837713, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 8.115789473684212e-06, | |
| "loss": 0.4244, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 1.901081916537867, | |
| "grad_norm": 9.125, | |
| "learning_rate": 8.105263157894736e-06, | |
| "loss": 0.3926, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.9026275115919629, | |
| "grad_norm": 10.75, | |
| "learning_rate": 8.094736842105264e-06, | |
| "loss": 0.4463, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 1.9041731066460588, | |
| "grad_norm": 13.375, | |
| "learning_rate": 8.08421052631579e-06, | |
| "loss": 0.4225, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 1.9057187017001547, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 8.073684210526317e-06, | |
| "loss": 0.4547, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 1.9072642967542504, | |
| "grad_norm": 10.5, | |
| "learning_rate": 8.063157894736843e-06, | |
| "loss": 0.427, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 1.9088098918083463, | |
| "grad_norm": 16.75, | |
| "learning_rate": 8.052631578947368e-06, | |
| "loss": 0.45, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.910355486862442, | |
| "grad_norm": 9.75, | |
| "learning_rate": 8.042105263157896e-06, | |
| "loss": 0.4475, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 1.9119010819165378, | |
| "grad_norm": 9.875, | |
| "learning_rate": 8.03157894736842e-06, | |
| "loss": 0.4848, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 1.9134466769706338, | |
| "grad_norm": 12.0, | |
| "learning_rate": 8.021052631578949e-06, | |
| "loss": 0.485, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 1.9149922720247297, | |
| "grad_norm": 10.5, | |
| "learning_rate": 8.010526315789473e-06, | |
| "loss": 0.4884, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 1.9165378670788253, | |
| "grad_norm": 10.5, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4178, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.918083462132921, | |
| "grad_norm": 11.25, | |
| "learning_rate": 7.989473684210528e-06, | |
| "loss": 0.4832, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 1.919629057187017, | |
| "grad_norm": 11.625, | |
| "learning_rate": 7.978947368421052e-06, | |
| "loss": 0.4736, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 1.9211746522411128, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 7.96842105263158e-06, | |
| "loss": 0.4225, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 1.9227202472952087, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 7.957894736842105e-06, | |
| "loss": 0.4381, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 1.9242658423493046, | |
| "grad_norm": 11.375, | |
| "learning_rate": 7.947368421052633e-06, | |
| "loss": 0.4732, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.9258114374034003, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.936842105263158e-06, | |
| "loss": 0.4806, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 1.927357032457496, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.926315789473686e-06, | |
| "loss": 0.371, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 1.928902627511592, | |
| "grad_norm": 9.75, | |
| "learning_rate": 7.915789473684212e-06, | |
| "loss": 0.4921, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 1.9304482225656878, | |
| "grad_norm": 14.8125, | |
| "learning_rate": 7.905263157894737e-06, | |
| "loss": 0.5108, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 1.9319938176197837, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 7.894736842105265e-06, | |
| "loss": 0.4646, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9335394126738794, | |
| "grad_norm": 10.0, | |
| "learning_rate": 7.88421052631579e-06, | |
| "loss": 0.4252, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 1.9350850077279753, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 7.873684210526317e-06, | |
| "loss": 0.4493, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 1.936630602782071, | |
| "grad_norm": 10.875, | |
| "learning_rate": 7.863157894736842e-06, | |
| "loss": 0.4008, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 1.938176197836167, | |
| "grad_norm": 15.5, | |
| "learning_rate": 7.85263157894737e-06, | |
| "loss": 0.4658, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 1.9397217928902628, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.842105263157895e-06, | |
| "loss": 0.4306, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 1.9412673879443587, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.831578947368421e-06, | |
| "loss": 0.5087, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 1.9428129829984544, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 7.821052631578949e-06, | |
| "loss": 0.4678, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 1.94435857805255, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 7.810526315789474e-06, | |
| "loss": 0.4189, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 1.945904173106646, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.4527, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 1.947449768160742, | |
| "grad_norm": 11.75, | |
| "learning_rate": 7.789473684210526e-06, | |
| "loss": 0.499, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9489953632148378, | |
| "grad_norm": 9.25, | |
| "learning_rate": 7.778947368421054e-06, | |
| "loss": 0.4036, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 1.9505409582689337, | |
| "grad_norm": 10.25, | |
| "learning_rate": 7.768421052631579e-06, | |
| "loss": 0.4725, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 1.9520865533230294, | |
| "grad_norm": 8.5, | |
| "learning_rate": 7.757894736842105e-06, | |
| "loss": 0.4144, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 1.953632148377125, | |
| "grad_norm": 13.0, | |
| "learning_rate": 7.747368421052631e-06, | |
| "loss": 0.4833, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 1.955177743431221, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 7.736842105263158e-06, | |
| "loss": 0.4006, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 1.9567233384853169, | |
| "grad_norm": 10.875, | |
| "learning_rate": 7.726315789473686e-06, | |
| "loss": 0.4265, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 1.9582689335394128, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 7.71578947368421e-06, | |
| "loss": 0.4581, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 1.9598145285935085, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 7.705263157894738e-06, | |
| "loss": 0.4954, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 1.9613601236476044, | |
| "grad_norm": 9.875, | |
| "learning_rate": 7.694736842105263e-06, | |
| "loss": 0.4877, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 1.9629057187017, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 7.68421052631579e-06, | |
| "loss": 0.4149, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.964451313755796, | |
| "grad_norm": 10.125, | |
| "learning_rate": 7.673684210526316e-06, | |
| "loss": 0.4674, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 1.9659969088098919, | |
| "grad_norm": 11.0, | |
| "learning_rate": 7.663157894736842e-06, | |
| "loss": 0.447, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 1.9675425038639878, | |
| "grad_norm": 11.5, | |
| "learning_rate": 7.65263157894737e-06, | |
| "loss": 0.4988, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 1.9690880989180835, | |
| "grad_norm": 12.625, | |
| "learning_rate": 7.642105263157895e-06, | |
| "loss": 0.4138, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 1.9706336939721791, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 7.631578947368423e-06, | |
| "loss": 0.4018, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.972179289026275, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 7.621052631578948e-06, | |
| "loss": 0.4473, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 1.973724884080371, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 7.610526315789474e-06, | |
| "loss": 0.4526, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 1.9752704791344669, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.4457, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 1.9768160741885628, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 7.589473684210526e-06, | |
| "loss": 0.4082, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 1.9783616692426584, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 7.578947368421054e-06, | |
| "loss": 0.4891, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9799072642967541, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 7.568421052631579e-06, | |
| "loss": 0.4421, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 1.98145285935085, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 7.557894736842106e-06, | |
| "loss": 0.4449, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 1.982998454404946, | |
| "grad_norm": 11.875, | |
| "learning_rate": 7.547368421052632e-06, | |
| "loss": 0.5123, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 1.9845440494590418, | |
| "grad_norm": 9.625, | |
| "learning_rate": 7.536842105263158e-06, | |
| "loss": 0.4362, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 1.9860896445131375, | |
| "grad_norm": 12.375, | |
| "learning_rate": 7.526315789473685e-06, | |
| "loss": 0.446, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 1.9876352395672334, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 7.515789473684211e-06, | |
| "loss": 0.4251, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 1.989180834621329, | |
| "grad_norm": 11.875, | |
| "learning_rate": 7.505263157894738e-06, | |
| "loss": 0.4227, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 1.990726429675425, | |
| "grad_norm": 11.75, | |
| "learning_rate": 7.494736842105263e-06, | |
| "loss": 0.4794, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 1.992272024729521, | |
| "grad_norm": 10.625, | |
| "learning_rate": 7.4842105263157905e-06, | |
| "loss": 0.4814, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 1.9938176197836168, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 7.473684210526316e-06, | |
| "loss": 0.421, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.9953632148377125, | |
| "grad_norm": 10.125, | |
| "learning_rate": 7.463157894736843e-06, | |
| "loss": 0.4335, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 1.9969088098918082, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 7.4526315789473695e-06, | |
| "loss": 0.4998, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 1.998454404945904, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 7.442105263157895e-06, | |
| "loss": 0.3654, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 9.5, | |
| "learning_rate": 7.431578947368422e-06, | |
| "loss": 0.421, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 2.001545595054096, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.421052631578948e-06, | |
| "loss": 0.3409, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.003091190108192, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 7.410526315789475e-06, | |
| "loss": 0.4153, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 2.0046367851622873, | |
| "grad_norm": 8.0, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.4277, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 2.006182380216383, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 7.3894736842105275e-06, | |
| "loss": 0.4332, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 2.007727975270479, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.378947368421053e-06, | |
| "loss": 0.3774, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 2.009273570324575, | |
| "grad_norm": 9.375, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 0.4583, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.010819165378671, | |
| "grad_norm": 9.75, | |
| "learning_rate": 7.3578947368421065e-06, | |
| "loss": 0.4037, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 2.012364760432767, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 7.347368421052632e-06, | |
| "loss": 0.4613, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 2.0139103554868623, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 7.336842105263159e-06, | |
| "loss": 0.45, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 2.015455950540958, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 7.326315789473685e-06, | |
| "loss": 0.478, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 2.017001545595054, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 7.315789473684212e-06, | |
| "loss": 0.4099, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.01854714064915, | |
| "grad_norm": 17.5, | |
| "learning_rate": 7.305263157894737e-06, | |
| "loss": 0.4187, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 2.020092735703246, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 7.2947368421052636e-06, | |
| "loss": 0.429, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 2.021638330757342, | |
| "grad_norm": 12.375, | |
| "learning_rate": 7.28421052631579e-06, | |
| "loss": 0.4693, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 2.0231839258114372, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 7.273684210526316e-06, | |
| "loss": 0.4918, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 2.024729520865533, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 7.263157894736843e-06, | |
| "loss": 0.4465, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.026275115919629, | |
| "grad_norm": 10.625, | |
| "learning_rate": 7.252631578947369e-06, | |
| "loss": 0.4254, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 2.027820710973725, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.242105263157896e-06, | |
| "loss": 0.4116, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 2.029366306027821, | |
| "grad_norm": 37.75, | |
| "learning_rate": 7.2315789473684215e-06, | |
| "loss": 0.4356, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 2.0309119010819163, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.221052631578948e-06, | |
| "loss": 0.4526, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 2.0324574961360122, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 7.210526315789474e-06, | |
| "loss": 0.4862, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.034003091190108, | |
| "grad_norm": 8.75, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.3943, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 2.035548686244204, | |
| "grad_norm": 8.625, | |
| "learning_rate": 7.189473684210527e-06, | |
| "loss": 0.3862, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 2.0370942812983, | |
| "grad_norm": 10.0, | |
| "learning_rate": 7.178947368421053e-06, | |
| "loss": 0.3895, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 2.038639876352396, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 7.16842105263158e-06, | |
| "loss": 0.3919, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 2.0401854714064913, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.157894736842106e-06, | |
| "loss": 0.4206, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.041731066460587, | |
| "grad_norm": 9.125, | |
| "learning_rate": 7.147368421052631e-06, | |
| "loss": 0.4216, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 2.043276661514683, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 7.1368421052631585e-06, | |
| "loss": 0.3875, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 2.044822256568779, | |
| "grad_norm": 8.625, | |
| "learning_rate": 7.126315789473685e-06, | |
| "loss": 0.394, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 2.046367851622875, | |
| "grad_norm": 10.0, | |
| "learning_rate": 7.115789473684211e-06, | |
| "loss": 0.398, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 2.047913446676971, | |
| "grad_norm": 9.5, | |
| "learning_rate": 7.1052631578947375e-06, | |
| "loss": 0.3825, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.0494590417310663, | |
| "grad_norm": 9.625, | |
| "learning_rate": 7.094736842105265e-06, | |
| "loss": 0.4619, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 2.051004636785162, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 7.08421052631579e-06, | |
| "loss": 0.3744, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 2.052550231839258, | |
| "grad_norm": 8.625, | |
| "learning_rate": 7.073684210526316e-06, | |
| "loss": 0.3775, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 2.054095826893354, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 7.063157894736843e-06, | |
| "loss": 0.4167, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 2.05564142194745, | |
| "grad_norm": 11.625, | |
| "learning_rate": 7.052631578947369e-06, | |
| "loss": 0.3926, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0571870170015454, | |
| "grad_norm": 14.25, | |
| "learning_rate": 7.0421052631578954e-06, | |
| "loss": 0.4458, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 2.0587326120556413, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.031578947368422e-06, | |
| "loss": 0.3743, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 2.060278207109737, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 7.021052631578948e-06, | |
| "loss": 0.4417, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 2.061823802163833, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 7.010526315789474e-06, | |
| "loss": 0.4122, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 2.063369397217929, | |
| "grad_norm": 9.625, | |
| "learning_rate": 7e-06, | |
| "loss": 0.4043, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.064914992272025, | |
| "grad_norm": 9.625, | |
| "learning_rate": 6.989473684210527e-06, | |
| "loss": 0.3708, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 2.0664605873261204, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 6.9789473684210525e-06, | |
| "loss": 0.4429, | |
| "step": 1337 | |
| }, | |
| { | |
| "epoch": 2.0680061823802163, | |
| "grad_norm": 10.25, | |
| "learning_rate": 6.96842105263158e-06, | |
| "loss": 0.392, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 2.069551777434312, | |
| "grad_norm": 11.375, | |
| "learning_rate": 6.957894736842106e-06, | |
| "loss": 0.3632, | |
| "step": 1339 | |
| }, | |
| { | |
| "epoch": 2.071097372488408, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 6.947368421052632e-06, | |
| "loss": 0.4258, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.072642967542504, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 6.936842105263159e-06, | |
| "loss": 0.4532, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 2.0741885625966, | |
| "grad_norm": 10.75, | |
| "learning_rate": 6.926315789473684e-06, | |
| "loss": 0.3734, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 2.0757341576506954, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 6.915789473684211e-06, | |
| "loss": 0.4601, | |
| "step": 1343 | |
| }, | |
| { | |
| "epoch": 2.0772797527047913, | |
| "grad_norm": 10.125, | |
| "learning_rate": 6.905263157894737e-06, | |
| "loss": 0.3973, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 2.078825347758887, | |
| "grad_norm": 10.0, | |
| "learning_rate": 6.894736842105264e-06, | |
| "loss": 0.3799, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.080370942812983, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 6.8842105263157895e-06, | |
| "loss": 0.3657, | |
| "step": 1346 | |
| }, | |
| { | |
| "epoch": 2.081916537867079, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 6.873684210526317e-06, | |
| "loss": 0.3858, | |
| "step": 1347 | |
| }, | |
| { | |
| "epoch": 2.0834621329211744, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 6.863157894736843e-06, | |
| "loss": 0.4277, | |
| "step": 1348 | |
| }, | |
| { | |
| "epoch": 2.0850077279752703, | |
| "grad_norm": 14.125, | |
| "learning_rate": 6.8526315789473685e-06, | |
| "loss": 0.3472, | |
| "step": 1349 | |
| }, | |
| { | |
| "epoch": 2.0865533230293662, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 6.842105263157896e-06, | |
| "loss": 0.3654, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.088098918083462, | |
| "grad_norm": 8.75, | |
| "learning_rate": 6.831578947368421e-06, | |
| "loss": 0.3987, | |
| "step": 1351 | |
| }, | |
| { | |
| "epoch": 2.089644513137558, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 6.821052631578948e-06, | |
| "loss": 0.3386, | |
| "step": 1352 | |
| }, | |
| { | |
| "epoch": 2.091190108191654, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 6.810526315789474e-06, | |
| "loss": 0.466, | |
| "step": 1353 | |
| }, | |
| { | |
| "epoch": 2.0927357032457494, | |
| "grad_norm": 10.0, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.4296, | |
| "step": 1354 | |
| }, | |
| { | |
| "epoch": 2.0942812982998453, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 6.789473684210527e-06, | |
| "loss": 0.3884, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.0958268933539412, | |
| "grad_norm": 10.75, | |
| "learning_rate": 6.778947368421053e-06, | |
| "loss": 0.3552, | |
| "step": 1356 | |
| }, | |
| { | |
| "epoch": 2.097372488408037, | |
| "grad_norm": 10.375, | |
| "learning_rate": 6.76842105263158e-06, | |
| "loss": 0.4134, | |
| "step": 1357 | |
| }, | |
| { | |
| "epoch": 2.098918083462133, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 6.7578947368421054e-06, | |
| "loss": 0.429, | |
| "step": 1358 | |
| }, | |
| { | |
| "epoch": 2.100463678516229, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 6.747368421052633e-06, | |
| "loss": 0.3613, | |
| "step": 1359 | |
| }, | |
| { | |
| "epoch": 2.1020092735703244, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 6.736842105263158e-06, | |
| "loss": 0.3704, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1035548686244203, | |
| "grad_norm": 11.625, | |
| "learning_rate": 6.726315789473685e-06, | |
| "loss": 0.4295, | |
| "step": 1361 | |
| }, | |
| { | |
| "epoch": 2.105100463678516, | |
| "grad_norm": 9.75, | |
| "learning_rate": 6.715789473684211e-06, | |
| "loss": 0.4071, | |
| "step": 1362 | |
| }, | |
| { | |
| "epoch": 2.106646058732612, | |
| "grad_norm": 10.0, | |
| "learning_rate": 6.705263157894737e-06, | |
| "loss": 0.447, | |
| "step": 1363 | |
| }, | |
| { | |
| "epoch": 2.108191653786708, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 6.694736842105264e-06, | |
| "loss": 0.4299, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 2.109737248840804, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 6.68421052631579e-06, | |
| "loss": 0.3642, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.1112828438948994, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 6.673684210526317e-06, | |
| "loss": 0.4116, | |
| "step": 1366 | |
| }, | |
| { | |
| "epoch": 2.1128284389489953, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 6.663157894736842e-06, | |
| "loss": 0.4034, | |
| "step": 1367 | |
| }, | |
| { | |
| "epoch": 2.114374034003091, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 6.6526315789473695e-06, | |
| "loss": 0.3993, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 2.115919629057187, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 6.642105263157895e-06, | |
| "loss": 0.4173, | |
| "step": 1369 | |
| }, | |
| { | |
| "epoch": 2.117465224111283, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 6.631578947368421e-06, | |
| "loss": 0.4035, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1190108191653785, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 6.621052631578948e-06, | |
| "loss": 0.383, | |
| "step": 1371 | |
| }, | |
| { | |
| "epoch": 2.1205564142194744, | |
| "grad_norm": 9.125, | |
| "learning_rate": 6.610526315789474e-06, | |
| "loss": 0.3614, | |
| "step": 1372 | |
| }, | |
| { | |
| "epoch": 2.1221020092735703, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 0.4334, | |
| "step": 1373 | |
| }, | |
| { | |
| "epoch": 2.123647604327666, | |
| "grad_norm": 9.5, | |
| "learning_rate": 6.589473684210527e-06, | |
| "loss": 0.4083, | |
| "step": 1374 | |
| }, | |
| { | |
| "epoch": 2.125193199381762, | |
| "grad_norm": 9.0, | |
| "learning_rate": 6.578947368421054e-06, | |
| "loss": 0.4057, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.126738794435858, | |
| "grad_norm": 9.75, | |
| "learning_rate": 6.568421052631579e-06, | |
| "loss": 0.3865, | |
| "step": 1376 | |
| }, | |
| { | |
| "epoch": 2.1282843894899535, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 6.557894736842106e-06, | |
| "loss": 0.3658, | |
| "step": 1377 | |
| }, | |
| { | |
| "epoch": 2.1298299845440494, | |
| "grad_norm": 9.25, | |
| "learning_rate": 6.547368421052632e-06, | |
| "loss": 0.379, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 2.1313755795981453, | |
| "grad_norm": 9.0, | |
| "learning_rate": 6.536842105263158e-06, | |
| "loss": 0.3727, | |
| "step": 1379 | |
| }, | |
| { | |
| "epoch": 2.132921174652241, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 6.526315789473685e-06, | |
| "loss": 0.4024, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.134466769706337, | |
| "grad_norm": 9.375, | |
| "learning_rate": 6.515789473684211e-06, | |
| "loss": 0.372, | |
| "step": 1381 | |
| }, | |
| { | |
| "epoch": 2.1360123647604325, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 6.505263157894738e-06, | |
| "loss": 0.3966, | |
| "step": 1382 | |
| }, | |
| { | |
| "epoch": 2.1375579598145285, | |
| "grad_norm": 18.5, | |
| "learning_rate": 6.494736842105264e-06, | |
| "loss": 0.396, | |
| "step": 1383 | |
| }, | |
| { | |
| "epoch": 2.1391035548686244, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 6.484210526315789e-06, | |
| "loss": 0.4434, | |
| "step": 1384 | |
| }, | |
| { | |
| "epoch": 2.1406491499227203, | |
| "grad_norm": 9.875, | |
| "learning_rate": 6.473684210526316e-06, | |
| "loss": 0.4041, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.142194744976816, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 6.463157894736843e-06, | |
| "loss": 0.3848, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 2.143740340030912, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 6.452631578947369e-06, | |
| "loss": 0.4139, | |
| "step": 1387 | |
| }, | |
| { | |
| "epoch": 2.1452859350850075, | |
| "grad_norm": 13.375, | |
| "learning_rate": 6.442105263157895e-06, | |
| "loss": 0.42, | |
| "step": 1388 | |
| }, | |
| { | |
| "epoch": 2.1468315301391034, | |
| "grad_norm": 10.75, | |
| "learning_rate": 6.431578947368422e-06, | |
| "loss": 0.4103, | |
| "step": 1389 | |
| }, | |
| { | |
| "epoch": 2.1483771251931993, | |
| "grad_norm": 20.125, | |
| "learning_rate": 6.421052631578948e-06, | |
| "loss": 0.4167, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1499227202472952, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 6.410526315789473e-06, | |
| "loss": 0.3943, | |
| "step": 1391 | |
| }, | |
| { | |
| "epoch": 2.151468315301391, | |
| "grad_norm": 10.0, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.4198, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 2.153013910355487, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 6.389473684210527e-06, | |
| "loss": 0.3931, | |
| "step": 1393 | |
| }, | |
| { | |
| "epoch": 2.1545595054095825, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 6.378947368421053e-06, | |
| "loss": 0.4473, | |
| "step": 1394 | |
| }, | |
| { | |
| "epoch": 2.1561051004636784, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 6.3684210526315795e-06, | |
| "loss": 0.4545, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.1576506955177743, | |
| "grad_norm": 10.625, | |
| "learning_rate": 6.357894736842106e-06, | |
| "loss": 0.4511, | |
| "step": 1396 | |
| }, | |
| { | |
| "epoch": 2.1591962905718702, | |
| "grad_norm": 9.5, | |
| "learning_rate": 6.347368421052632e-06, | |
| "loss": 0.3953, | |
| "step": 1397 | |
| }, | |
| { | |
| "epoch": 2.160741885625966, | |
| "grad_norm": 9.5, | |
| "learning_rate": 6.336842105263158e-06, | |
| "loss": 0.4376, | |
| "step": 1398 | |
| }, | |
| { | |
| "epoch": 2.162287480680062, | |
| "grad_norm": 10.125, | |
| "learning_rate": 6.326315789473685e-06, | |
| "loss": 0.3919, | |
| "step": 1399 | |
| }, | |
| { | |
| "epoch": 2.1638330757341575, | |
| "grad_norm": 11.625, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 0.4433, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1653786707882534, | |
| "grad_norm": 10.75, | |
| "learning_rate": 6.3052631578947375e-06, | |
| "loss": 0.4774, | |
| "step": 1401 | |
| }, | |
| { | |
| "epoch": 2.1669242658423493, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 6.294736842105264e-06, | |
| "loss": 0.3942, | |
| "step": 1402 | |
| }, | |
| { | |
| "epoch": 2.1684698608964452, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 6.28421052631579e-06, | |
| "loss": 0.4453, | |
| "step": 1403 | |
| }, | |
| { | |
| "epoch": 2.170015455950541, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 6.2736842105263165e-06, | |
| "loss": 0.3921, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 2.1715610510046366, | |
| "grad_norm": 11.625, | |
| "learning_rate": 6.263157894736842e-06, | |
| "loss": 0.405, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 2.1731066460587325, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 6.252631578947369e-06, | |
| "loss": 0.3684, | |
| "step": 1406 | |
| }, | |
| { | |
| "epoch": 2.1746522411128284, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 6.242105263157895e-06, | |
| "loss": 0.4626, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 2.1761978361669243, | |
| "grad_norm": 10.25, | |
| "learning_rate": 6.231578947368422e-06, | |
| "loss": 0.422, | |
| "step": 1408 | |
| }, | |
| { | |
| "epoch": 2.17774343122102, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 6.221052631578947e-06, | |
| "loss": 0.4048, | |
| "step": 1409 | |
| }, | |
| { | |
| "epoch": 2.179289026275116, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 6.2105263157894745e-06, | |
| "loss": 0.4063, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1808346213292116, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 0.4196, | |
| "step": 1411 | |
| }, | |
| { | |
| "epoch": 2.1823802163833075, | |
| "grad_norm": 9.0, | |
| "learning_rate": 6.189473684210526e-06, | |
| "loss": 0.3823, | |
| "step": 1412 | |
| }, | |
| { | |
| "epoch": 2.1839258114374034, | |
| "grad_norm": 9.75, | |
| "learning_rate": 6.1789473684210534e-06, | |
| "loss": 0.4392, | |
| "step": 1413 | |
| }, | |
| { | |
| "epoch": 2.1854714064914993, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 6.168421052631579e-06, | |
| "loss": 0.4116, | |
| "step": 1414 | |
| }, | |
| { | |
| "epoch": 2.187017001545595, | |
| "grad_norm": 9.0, | |
| "learning_rate": 6.157894736842106e-06, | |
| "loss": 0.4159, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 2.1885625965996907, | |
| "grad_norm": 11.125, | |
| "learning_rate": 6.1473684210526316e-06, | |
| "loss": 0.3939, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 2.1901081916537866, | |
| "grad_norm": 10.5, | |
| "learning_rate": 6.136842105263159e-06, | |
| "loss": 0.3893, | |
| "step": 1417 | |
| }, | |
| { | |
| "epoch": 2.1916537867078825, | |
| "grad_norm": 9.25, | |
| "learning_rate": 6.126315789473685e-06, | |
| "loss": 0.3817, | |
| "step": 1418 | |
| }, | |
| { | |
| "epoch": 2.1931993817619784, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 6.1157894736842106e-06, | |
| "loss": 0.4388, | |
| "step": 1419 | |
| }, | |
| { | |
| "epoch": 2.1947449768160743, | |
| "grad_norm": 12.125, | |
| "learning_rate": 6.105263157894738e-06, | |
| "loss": 0.5682, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.19629057187017, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 6.094736842105263e-06, | |
| "loss": 0.3891, | |
| "step": 1421 | |
| }, | |
| { | |
| "epoch": 2.1978361669242656, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 6.08421052631579e-06, | |
| "loss": 0.3766, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 2.1993817619783615, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 6.073684210526316e-06, | |
| "loss": 0.4692, | |
| "step": 1423 | |
| }, | |
| { | |
| "epoch": 2.2009273570324575, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 6.063157894736843e-06, | |
| "loss": 0.3734, | |
| "step": 1424 | |
| }, | |
| { | |
| "epoch": 2.2024729520865534, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 6.0526315789473685e-06, | |
| "loss": 0.3428, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.2040185471406493, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 6.042105263157895e-06, | |
| "loss": 0.4315, | |
| "step": 1426 | |
| }, | |
| { | |
| "epoch": 2.205564142194745, | |
| "grad_norm": 15.125, | |
| "learning_rate": 6.031578947368422e-06, | |
| "loss": 0.4135, | |
| "step": 1427 | |
| }, | |
| { | |
| "epoch": 2.2071097372488406, | |
| "grad_norm": 10.25, | |
| "learning_rate": 6.0210526315789475e-06, | |
| "loss": 0.4742, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 2.2086553323029365, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 6.010526315789475e-06, | |
| "loss": 0.3849, | |
| "step": 1429 | |
| }, | |
| { | |
| "epoch": 2.2102009273570324, | |
| "grad_norm": 10.0, | |
| "learning_rate": 6e-06, | |
| "loss": 0.4307, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2117465224111283, | |
| "grad_norm": 9.25, | |
| "learning_rate": 5.989473684210527e-06, | |
| "loss": 0.3806, | |
| "step": 1431 | |
| }, | |
| { | |
| "epoch": 2.2132921174652243, | |
| "grad_norm": 9.875, | |
| "learning_rate": 5.978947368421053e-06, | |
| "loss": 0.3976, | |
| "step": 1432 | |
| }, | |
| { | |
| "epoch": 2.21483771251932, | |
| "grad_norm": 10.125, | |
| "learning_rate": 5.968421052631579e-06, | |
| "loss": 0.4277, | |
| "step": 1433 | |
| }, | |
| { | |
| "epoch": 2.2163833075734156, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 5.9578947368421055e-06, | |
| "loss": 0.3941, | |
| "step": 1434 | |
| }, | |
| { | |
| "epoch": 2.2179289026275115, | |
| "grad_norm": 11.75, | |
| "learning_rate": 5.947368421052632e-06, | |
| "loss": 0.4576, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 2.2194744976816074, | |
| "grad_norm": 9.5, | |
| "learning_rate": 5.936842105263159e-06, | |
| "loss": 0.4679, | |
| "step": 1436 | |
| }, | |
| { | |
| "epoch": 2.2210200927357033, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 5.9263157894736844e-06, | |
| "loss": 0.4179, | |
| "step": 1437 | |
| }, | |
| { | |
| "epoch": 2.2225656877897992, | |
| "grad_norm": 10.75, | |
| "learning_rate": 5.915789473684212e-06, | |
| "loss": 0.4054, | |
| "step": 1438 | |
| }, | |
| { | |
| "epoch": 2.2241112828438947, | |
| "grad_norm": 9.375, | |
| "learning_rate": 5.905263157894737e-06, | |
| "loss": 0.4408, | |
| "step": 1439 | |
| }, | |
| { | |
| "epoch": 2.2256568778979906, | |
| "grad_norm": 8.75, | |
| "learning_rate": 5.8947368421052634e-06, | |
| "loss": 0.3918, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.2272024729520865, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 5.88421052631579e-06, | |
| "loss": 0.3736, | |
| "step": 1441 | |
| }, | |
| { | |
| "epoch": 2.2287480680061824, | |
| "grad_norm": 12.625, | |
| "learning_rate": 5.873684210526316e-06, | |
| "loss": 0.445, | |
| "step": 1442 | |
| }, | |
| { | |
| "epoch": 2.2302936630602783, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 5.863157894736842e-06, | |
| "loss": 0.4658, | |
| "step": 1443 | |
| }, | |
| { | |
| "epoch": 2.2318392581143742, | |
| "grad_norm": 9.375, | |
| "learning_rate": 5.852631578947369e-06, | |
| "loss": 0.3555, | |
| "step": 1444 | |
| }, | |
| { | |
| "epoch": 2.2333848531684697, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 5.842105263157896e-06, | |
| "loss": 0.4339, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 2.2349304482225656, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 5.831578947368421e-06, | |
| "loss": 0.4169, | |
| "step": 1446 | |
| }, | |
| { | |
| "epoch": 2.2364760432766615, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 5.8210526315789486e-06, | |
| "loss": 0.4089, | |
| "step": 1447 | |
| }, | |
| { | |
| "epoch": 2.2380216383307574, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 5.810526315789474e-06, | |
| "loss": 0.451, | |
| "step": 1448 | |
| }, | |
| { | |
| "epoch": 2.2395672333848533, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 5.8e-06, | |
| "loss": 0.4551, | |
| "step": 1449 | |
| }, | |
| { | |
| "epoch": 2.2411128284389488, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 5.789473684210527e-06, | |
| "loss": 0.405, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2426584234930447, | |
| "grad_norm": 9.875, | |
| "learning_rate": 5.778947368421053e-06, | |
| "loss": 0.4629, | |
| "step": 1451 | |
| }, | |
| { | |
| "epoch": 2.2442040185471406, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 5.76842105263158e-06, | |
| "loss": 0.3815, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 2.2457496136012365, | |
| "grad_norm": 10.375, | |
| "learning_rate": 5.757894736842106e-06, | |
| "loss": 0.4429, | |
| "step": 1453 | |
| }, | |
| { | |
| "epoch": 2.2472952086553324, | |
| "grad_norm": 10.125, | |
| "learning_rate": 5.747368421052633e-06, | |
| "loss": 0.4238, | |
| "step": 1454 | |
| }, | |
| { | |
| "epoch": 2.2488408037094283, | |
| "grad_norm": 10.125, | |
| "learning_rate": 5.736842105263158e-06, | |
| "loss": 0.3564, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 2.250386398763524, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 5.726315789473685e-06, | |
| "loss": 0.4415, | |
| "step": 1456 | |
| }, | |
| { | |
| "epoch": 2.2519319938176197, | |
| "grad_norm": 9.125, | |
| "learning_rate": 5.715789473684211e-06, | |
| "loss": 0.3871, | |
| "step": 1457 | |
| }, | |
| { | |
| "epoch": 2.2534775888717156, | |
| "grad_norm": 10.125, | |
| "learning_rate": 5.705263157894737e-06, | |
| "loss": 0.4082, | |
| "step": 1458 | |
| }, | |
| { | |
| "epoch": 2.2550231839258115, | |
| "grad_norm": 11.375, | |
| "learning_rate": 5.694736842105264e-06, | |
| "loss": 0.5043, | |
| "step": 1459 | |
| }, | |
| { | |
| "epoch": 2.2565687789799074, | |
| "grad_norm": 9.25, | |
| "learning_rate": 5.68421052631579e-06, | |
| "loss": 0.414, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2581143740340033, | |
| "grad_norm": 11.25, | |
| "learning_rate": 5.673684210526317e-06, | |
| "loss": 0.4273, | |
| "step": 1461 | |
| }, | |
| { | |
| "epoch": 2.2596599690880987, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 5.663157894736843e-06, | |
| "loss": 0.366, | |
| "step": 1462 | |
| }, | |
| { | |
| "epoch": 2.2612055641421946, | |
| "grad_norm": 28.375, | |
| "learning_rate": 5.652631578947368e-06, | |
| "loss": 0.4265, | |
| "step": 1463 | |
| }, | |
| { | |
| "epoch": 2.2627511591962906, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 5.642105263157895e-06, | |
| "loss": 0.417, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 2.2642967542503865, | |
| "grad_norm": 8.875, | |
| "learning_rate": 5.631578947368422e-06, | |
| "loss": 0.4107, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 2.2658423493044824, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 5.621052631578948e-06, | |
| "loss": 0.4519, | |
| "step": 1466 | |
| }, | |
| { | |
| "epoch": 2.2673879443585783, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 5.610526315789474e-06, | |
| "loss": 0.4128, | |
| "step": 1467 | |
| }, | |
| { | |
| "epoch": 2.2689335394126737, | |
| "grad_norm": 10.75, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.4795, | |
| "step": 1468 | |
| }, | |
| { | |
| "epoch": 2.2704791344667696, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 5.589473684210527e-06, | |
| "loss": 0.3894, | |
| "step": 1469 | |
| }, | |
| { | |
| "epoch": 2.2720247295208655, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 5.578947368421052e-06, | |
| "loss": 0.4256, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2735703245749614, | |
| "grad_norm": 10.75, | |
| "learning_rate": 5.5684210526315796e-06, | |
| "loss": 0.397, | |
| "step": 1471 | |
| }, | |
| { | |
| "epoch": 2.2751159196290573, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 5.557894736842105e-06, | |
| "loss": 0.4466, | |
| "step": 1472 | |
| }, | |
| { | |
| "epoch": 2.276661514683153, | |
| "grad_norm": 12.125, | |
| "learning_rate": 5.547368421052632e-06, | |
| "loss": 0.4147, | |
| "step": 1473 | |
| }, | |
| { | |
| "epoch": 2.2782071097372487, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 5.5368421052631586e-06, | |
| "loss": 0.3968, | |
| "step": 1474 | |
| }, | |
| { | |
| "epoch": 2.2797527047913446, | |
| "grad_norm": 21.375, | |
| "learning_rate": 5.526315789473685e-06, | |
| "loss": 0.4401, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.2812982998454405, | |
| "grad_norm": 11.75, | |
| "learning_rate": 5.515789473684211e-06, | |
| "loss": 0.3951, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 2.2828438948995364, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 5.505263157894737e-06, | |
| "loss": 0.3654, | |
| "step": 1477 | |
| }, | |
| { | |
| "epoch": 2.2843894899536323, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 5.494736842105264e-06, | |
| "loss": 0.4121, | |
| "step": 1478 | |
| }, | |
| { | |
| "epoch": 2.285935085007728, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 5.484210526315789e-06, | |
| "loss": 0.4137, | |
| "step": 1479 | |
| }, | |
| { | |
| "epoch": 2.2874806800618237, | |
| "grad_norm": 10.0, | |
| "learning_rate": 5.4736842105263165e-06, | |
| "loss": 0.508, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.2890262751159196, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 5.463157894736843e-06, | |
| "loss": 0.436, | |
| "step": 1481 | |
| }, | |
| { | |
| "epoch": 2.2905718701700155, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 5.452631578947369e-06, | |
| "loss": 0.4118, | |
| "step": 1482 | |
| }, | |
| { | |
| "epoch": 2.2921174652241114, | |
| "grad_norm": 11.375, | |
| "learning_rate": 5.4421052631578955e-06, | |
| "loss": 0.4587, | |
| "step": 1483 | |
| }, | |
| { | |
| "epoch": 2.293663060278207, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 5.431578947368421e-06, | |
| "loss": 0.3737, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 2.295208655332303, | |
| "grad_norm": 10.25, | |
| "learning_rate": 5.421052631578948e-06, | |
| "loss": 0.4325, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 2.2967542503863987, | |
| "grad_norm": 10.25, | |
| "learning_rate": 5.410526315789474e-06, | |
| "loss": 0.4284, | |
| "step": 1486 | |
| }, | |
| { | |
| "epoch": 2.2982998454404946, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 0.4085, | |
| "step": 1487 | |
| }, | |
| { | |
| "epoch": 2.2998454404945905, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 5.389473684210526e-06, | |
| "loss": 0.374, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 2.3013910355486864, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 5.3789473684210535e-06, | |
| "loss": 0.3844, | |
| "step": 1489 | |
| }, | |
| { | |
| "epoch": 2.3029366306027823, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 5.36842105263158e-06, | |
| "loss": 0.4483, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.3044822256568778, | |
| "grad_norm": 9.375, | |
| "learning_rate": 5.357894736842105e-06, | |
| "loss": 0.3997, | |
| "step": 1491 | |
| }, | |
| { | |
| "epoch": 2.3060278207109737, | |
| "grad_norm": 9.5, | |
| "learning_rate": 5.3473684210526325e-06, | |
| "loss": 0.3746, | |
| "step": 1492 | |
| }, | |
| { | |
| "epoch": 2.3075734157650696, | |
| "grad_norm": 11.375, | |
| "learning_rate": 5.336842105263158e-06, | |
| "loss": 0.4513, | |
| "step": 1493 | |
| }, | |
| { | |
| "epoch": 2.3091190108191655, | |
| "grad_norm": 12.125, | |
| "learning_rate": 5.326315789473685e-06, | |
| "loss": 0.4056, | |
| "step": 1494 | |
| }, | |
| { | |
| "epoch": 2.3106646058732614, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 5.315789473684211e-06, | |
| "loss": 0.3634, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 2.312210200927357, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 5.305263157894738e-06, | |
| "loss": 0.4042, | |
| "step": 1496 | |
| }, | |
| { | |
| "epoch": 2.3137557959814528, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 5.294736842105263e-06, | |
| "loss": 0.3689, | |
| "step": 1497 | |
| }, | |
| { | |
| "epoch": 2.3153013910355487, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 5.2842105263157896e-06, | |
| "loss": 0.3948, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 2.3168469860896446, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 5.273684210526317e-06, | |
| "loss": 0.4147, | |
| "step": 1499 | |
| }, | |
| { | |
| "epoch": 2.3183925811437405, | |
| "grad_norm": 33.0, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 0.4441, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.3199381761978364, | |
| "grad_norm": 10.5, | |
| "learning_rate": 5.252631578947369e-06, | |
| "loss": 0.4294, | |
| "step": 1501 | |
| }, | |
| { | |
| "epoch": 2.321483771251932, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 5.242105263157895e-06, | |
| "loss": 0.364, | |
| "step": 1502 | |
| }, | |
| { | |
| "epoch": 2.3230293663060277, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 5.231578947368422e-06, | |
| "loss": 0.4008, | |
| "step": 1503 | |
| }, | |
| { | |
| "epoch": 2.3245749613601236, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 5.2210526315789475e-06, | |
| "loss": 0.4188, | |
| "step": 1504 | |
| }, | |
| { | |
| "epoch": 2.3261205564142196, | |
| "grad_norm": 22.375, | |
| "learning_rate": 5.210526315789474e-06, | |
| "loss": 0.4373, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 2.3276661514683155, | |
| "grad_norm": 12.0, | |
| "learning_rate": 5.2e-06, | |
| "loss": 0.4485, | |
| "step": 1506 | |
| }, | |
| { | |
| "epoch": 2.329211746522411, | |
| "grad_norm": 11.125, | |
| "learning_rate": 5.1894736842105265e-06, | |
| "loss": 0.3989, | |
| "step": 1507 | |
| }, | |
| { | |
| "epoch": 2.330757341576507, | |
| "grad_norm": 11.0, | |
| "learning_rate": 5.178947368421054e-06, | |
| "loss": 0.4409, | |
| "step": 1508 | |
| }, | |
| { | |
| "epoch": 2.3323029366306027, | |
| "grad_norm": 10.625, | |
| "learning_rate": 5.168421052631579e-06, | |
| "loss": 0.4522, | |
| "step": 1509 | |
| }, | |
| { | |
| "epoch": 2.3338485316846986, | |
| "grad_norm": 9.875, | |
| "learning_rate": 5.157894736842106e-06, | |
| "loss": 0.4109, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.3353941267387945, | |
| "grad_norm": 9.25, | |
| "learning_rate": 5.147368421052632e-06, | |
| "loss": 0.3738, | |
| "step": 1511 | |
| }, | |
| { | |
| "epoch": 2.3369397217928904, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 5.136842105263158e-06, | |
| "loss": 0.4172, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 2.338485316846986, | |
| "grad_norm": 9.625, | |
| "learning_rate": 5.1263157894736845e-06, | |
| "loss": 0.4146, | |
| "step": 1513 | |
| }, | |
| { | |
| "epoch": 2.340030911901082, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 5.115789473684211e-06, | |
| "loss": 0.3463, | |
| "step": 1514 | |
| }, | |
| { | |
| "epoch": 2.3415765069551777, | |
| "grad_norm": 10.375, | |
| "learning_rate": 5.105263157894738e-06, | |
| "loss": 0.4149, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 2.3431221020092736, | |
| "grad_norm": 14.375, | |
| "learning_rate": 5.0947368421052635e-06, | |
| "loss": 0.4231, | |
| "step": 1516 | |
| }, | |
| { | |
| "epoch": 2.3446676970633695, | |
| "grad_norm": 10.125, | |
| "learning_rate": 5.084210526315791e-06, | |
| "loss": 0.4468, | |
| "step": 1517 | |
| }, | |
| { | |
| "epoch": 2.346213292117465, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 5.073684210526316e-06, | |
| "loss": 0.3793, | |
| "step": 1518 | |
| }, | |
| { | |
| "epoch": 2.347758887171561, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 5.0631578947368424e-06, | |
| "loss": 0.4346, | |
| "step": 1519 | |
| }, | |
| { | |
| "epoch": 2.349304482225657, | |
| "grad_norm": 10.0, | |
| "learning_rate": 5.052631578947369e-06, | |
| "loss": 0.3638, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3508500772797527, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 5.042105263157895e-06, | |
| "loss": 0.462, | |
| "step": 1521 | |
| }, | |
| { | |
| "epoch": 2.3523956723338486, | |
| "grad_norm": 10.0, | |
| "learning_rate": 5.0315789473684214e-06, | |
| "loss": 0.3755, | |
| "step": 1522 | |
| }, | |
| { | |
| "epoch": 2.3539412673879445, | |
| "grad_norm": 9.5, | |
| "learning_rate": 5.021052631578948e-06, | |
| "loss": 0.4176, | |
| "step": 1523 | |
| }, | |
| { | |
| "epoch": 2.3554868624420404, | |
| "grad_norm": 16.625, | |
| "learning_rate": 5.010526315789475e-06, | |
| "loss": 0.427, | |
| "step": 1524 | |
| }, | |
| { | |
| "epoch": 2.357032457496136, | |
| "grad_norm": 12.25, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4444, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 2.358578052550232, | |
| "grad_norm": 10.625, | |
| "learning_rate": 4.989473684210527e-06, | |
| "loss": 0.4319, | |
| "step": 1526 | |
| }, | |
| { | |
| "epoch": 2.3601236476043277, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 4.978947368421053e-06, | |
| "loss": 0.406, | |
| "step": 1527 | |
| }, | |
| { | |
| "epoch": 2.3616692426584236, | |
| "grad_norm": 12.125, | |
| "learning_rate": 4.968421052631579e-06, | |
| "loss": 0.4427, | |
| "step": 1528 | |
| }, | |
| { | |
| "epoch": 2.3632148377125195, | |
| "grad_norm": 18.25, | |
| "learning_rate": 4.957894736842106e-06, | |
| "loss": 0.4267, | |
| "step": 1529 | |
| }, | |
| { | |
| "epoch": 2.364760432766615, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 4.947368421052632e-06, | |
| "loss": 0.4236, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.366306027820711, | |
| "grad_norm": 13.75, | |
| "learning_rate": 4.936842105263158e-06, | |
| "loss": 0.3631, | |
| "step": 1531 | |
| }, | |
| { | |
| "epoch": 2.3678516228748068, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.926315789473685e-06, | |
| "loss": 0.4081, | |
| "step": 1532 | |
| }, | |
| { | |
| "epoch": 2.3693972179289027, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 4.915789473684211e-06, | |
| "loss": 0.431, | |
| "step": 1533 | |
| }, | |
| { | |
| "epoch": 2.3709428129829986, | |
| "grad_norm": 11.0, | |
| "learning_rate": 4.905263157894737e-06, | |
| "loss": 0.3904, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 2.3724884080370945, | |
| "grad_norm": 10.0, | |
| "learning_rate": 4.894736842105264e-06, | |
| "loss": 0.3249, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 2.37403400309119, | |
| "grad_norm": 10.25, | |
| "learning_rate": 4.88421052631579e-06, | |
| "loss": 0.4481, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 2.375579598145286, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 4.873684210526316e-06, | |
| "loss": 0.4045, | |
| "step": 1537 | |
| }, | |
| { | |
| "epoch": 2.3771251931993818, | |
| "grad_norm": 11.125, | |
| "learning_rate": 4.863157894736843e-06, | |
| "loss": 0.4266, | |
| "step": 1538 | |
| }, | |
| { | |
| "epoch": 2.3786707882534777, | |
| "grad_norm": 11.25, | |
| "learning_rate": 4.852631578947369e-06, | |
| "loss": 0.4105, | |
| "step": 1539 | |
| }, | |
| { | |
| "epoch": 2.3802163833075736, | |
| "grad_norm": 11.0, | |
| "learning_rate": 4.842105263157895e-06, | |
| "loss": 0.3739, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.381761978361669, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.831578947368422e-06, | |
| "loss": 0.367, | |
| "step": 1541 | |
| }, | |
| { | |
| "epoch": 2.383307573415765, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 4.821052631578948e-06, | |
| "loss": 0.4097, | |
| "step": 1542 | |
| }, | |
| { | |
| "epoch": 2.384853168469861, | |
| "grad_norm": 11.5, | |
| "learning_rate": 4.8105263157894735e-06, | |
| "loss": 0.3812, | |
| "step": 1543 | |
| }, | |
| { | |
| "epoch": 2.3863987635239567, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.3696, | |
| "step": 1544 | |
| }, | |
| { | |
| "epoch": 2.3879443585780527, | |
| "grad_norm": 16.375, | |
| "learning_rate": 4.789473684210527e-06, | |
| "loss": 0.4267, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 2.3894899536321486, | |
| "grad_norm": 11.125, | |
| "learning_rate": 4.778947368421053e-06, | |
| "loss": 0.445, | |
| "step": 1546 | |
| }, | |
| { | |
| "epoch": 2.391035548686244, | |
| "grad_norm": 9.625, | |
| "learning_rate": 4.76842105263158e-06, | |
| "loss": 0.4179, | |
| "step": 1547 | |
| }, | |
| { | |
| "epoch": 2.39258114374034, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 4.757894736842106e-06, | |
| "loss": 0.3801, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 2.394126738794436, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 4.747368421052632e-06, | |
| "loss": 0.3343, | |
| "step": 1549 | |
| }, | |
| { | |
| "epoch": 2.3956723338485317, | |
| "grad_norm": 24.875, | |
| "learning_rate": 4.736842105263158e-06, | |
| "loss": 0.4098, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.3972179289026276, | |
| "grad_norm": 11.75, | |
| "learning_rate": 4.726315789473684e-06, | |
| "loss": 0.4387, | |
| "step": 1551 | |
| }, | |
| { | |
| "epoch": 2.398763523956723, | |
| "grad_norm": 9.125, | |
| "learning_rate": 4.71578947368421e-06, | |
| "loss": 0.4161, | |
| "step": 1552 | |
| }, | |
| { | |
| "epoch": 2.400309119010819, | |
| "grad_norm": 11.25, | |
| "learning_rate": 4.705263157894738e-06, | |
| "loss": 0.4809, | |
| "step": 1553 | |
| }, | |
| { | |
| "epoch": 2.401854714064915, | |
| "grad_norm": 11.25, | |
| "learning_rate": 4.694736842105264e-06, | |
| "loss": 0.4091, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 2.403400309119011, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 4.68421052631579e-06, | |
| "loss": 0.3778, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 2.4049459041731067, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 4.6736842105263166e-06, | |
| "loss": 0.3738, | |
| "step": 1556 | |
| }, | |
| { | |
| "epoch": 2.4064914992272026, | |
| "grad_norm": 11.375, | |
| "learning_rate": 4.663157894736842e-06, | |
| "loss": 0.4912, | |
| "step": 1557 | |
| }, | |
| { | |
| "epoch": 2.4080370942812985, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.652631578947368e-06, | |
| "loss": 0.3971, | |
| "step": 1558 | |
| }, | |
| { | |
| "epoch": 2.409582689335394, | |
| "grad_norm": 11.625, | |
| "learning_rate": 4.642105263157895e-06, | |
| "loss": 0.3892, | |
| "step": 1559 | |
| }, | |
| { | |
| "epoch": 2.41112828438949, | |
| "grad_norm": 9.875, | |
| "learning_rate": 4.631578947368421e-06, | |
| "loss": 0.4003, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.412673879443586, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.621052631578948e-06, | |
| "loss": 0.4324, | |
| "step": 1561 | |
| }, | |
| { | |
| "epoch": 2.4142194744976817, | |
| "grad_norm": 9.375, | |
| "learning_rate": 4.6105263157894745e-06, | |
| "loss": 0.3877, | |
| "step": 1562 | |
| }, | |
| { | |
| "epoch": 2.4157650695517776, | |
| "grad_norm": 10.375, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.4197, | |
| "step": 1563 | |
| }, | |
| { | |
| "epoch": 2.417310664605873, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.589473684210526e-06, | |
| "loss": 0.3632, | |
| "step": 1564 | |
| }, | |
| { | |
| "epoch": 2.418856259659969, | |
| "grad_norm": 9.875, | |
| "learning_rate": 4.578947368421053e-06, | |
| "loss": 0.3915, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 2.420401854714065, | |
| "grad_norm": 8.625, | |
| "learning_rate": 4.568421052631579e-06, | |
| "loss": 0.3877, | |
| "step": 1566 | |
| }, | |
| { | |
| "epoch": 2.421947449768161, | |
| "grad_norm": 10.125, | |
| "learning_rate": 4.557894736842105e-06, | |
| "loss": 0.4359, | |
| "step": 1567 | |
| }, | |
| { | |
| "epoch": 2.4234930448222567, | |
| "grad_norm": 9.75, | |
| "learning_rate": 4.547368421052632e-06, | |
| "loss": 0.3725, | |
| "step": 1568 | |
| }, | |
| { | |
| "epoch": 2.4250386398763526, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 4.536842105263158e-06, | |
| "loss": 0.3291, | |
| "step": 1569 | |
| }, | |
| { | |
| "epoch": 2.426584234930448, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.526315789473685e-06, | |
| "loss": 0.3561, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.428129829984544, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.5157894736842115e-06, | |
| "loss": 0.4158, | |
| "step": 1571 | |
| }, | |
| { | |
| "epoch": 2.42967542503864, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 4.505263157894737e-06, | |
| "loss": 0.3719, | |
| "step": 1572 | |
| }, | |
| { | |
| "epoch": 2.4312210200927358, | |
| "grad_norm": 9.5, | |
| "learning_rate": 4.494736842105263e-06, | |
| "loss": 0.3725, | |
| "step": 1573 | |
| }, | |
| { | |
| "epoch": 2.4327666151468317, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 4.48421052631579e-06, | |
| "loss": 0.4742, | |
| "step": 1574 | |
| }, | |
| { | |
| "epoch": 2.434312210200927, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 4.473684210526316e-06, | |
| "loss": 0.4109, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 2.435857805255023, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 4.463157894736842e-06, | |
| "loss": 0.3921, | |
| "step": 1576 | |
| }, | |
| { | |
| "epoch": 2.437403400309119, | |
| "grad_norm": 10.625, | |
| "learning_rate": 4.452631578947369e-06, | |
| "loss": 0.4003, | |
| "step": 1577 | |
| }, | |
| { | |
| "epoch": 2.438948995363215, | |
| "grad_norm": 10.125, | |
| "learning_rate": 4.442105263157896e-06, | |
| "loss": 0.3946, | |
| "step": 1578 | |
| }, | |
| { | |
| "epoch": 2.4404945904173108, | |
| "grad_norm": 9.75, | |
| "learning_rate": 4.431578947368421e-06, | |
| "loss": 0.394, | |
| "step": 1579 | |
| }, | |
| { | |
| "epoch": 2.4420401854714067, | |
| "grad_norm": 10.875, | |
| "learning_rate": 4.4210526315789476e-06, | |
| "loss": 0.4385, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.443585780525502, | |
| "grad_norm": 11.5, | |
| "learning_rate": 4.410526315789474e-06, | |
| "loss": 0.4159, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 2.445131375579598, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.3802, | |
| "step": 1582 | |
| }, | |
| { | |
| "epoch": 2.446676970633694, | |
| "grad_norm": 10.375, | |
| "learning_rate": 4.3894736842105266e-06, | |
| "loss": 0.4597, | |
| "step": 1583 | |
| }, | |
| { | |
| "epoch": 2.44822256568779, | |
| "grad_norm": 10.25, | |
| "learning_rate": 4.378947368421053e-06, | |
| "loss": 0.4427, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 2.4497681607418857, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.368421052631579e-06, | |
| "loss": 0.4613, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 2.451313755795981, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.3578947368421055e-06, | |
| "loss": 0.4316, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 2.452859350850077, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 4.347368421052632e-06, | |
| "loss": 0.4197, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 2.454404945904173, | |
| "grad_norm": 14.75, | |
| "learning_rate": 4.336842105263158e-06, | |
| "loss": 0.411, | |
| "step": 1588 | |
| }, | |
| { | |
| "epoch": 2.455950540958269, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 4.3263157894736845e-06, | |
| "loss": 0.4625, | |
| "step": 1589 | |
| }, | |
| { | |
| "epoch": 2.457496136012365, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.315789473684211e-06, | |
| "loss": 0.418, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4590417310664607, | |
| "grad_norm": 20.125, | |
| "learning_rate": 4.305263157894737e-06, | |
| "loss": 0.379, | |
| "step": 1591 | |
| }, | |
| { | |
| "epoch": 2.4605873261205566, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 4.2947368421052635e-06, | |
| "loss": 0.3886, | |
| "step": 1592 | |
| }, | |
| { | |
| "epoch": 2.462132921174652, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 4.28421052631579e-06, | |
| "loss": 0.4662, | |
| "step": 1593 | |
| }, | |
| { | |
| "epoch": 2.463678516228748, | |
| "grad_norm": 12.75, | |
| "learning_rate": 4.273684210526316e-06, | |
| "loss": 0.4249, | |
| "step": 1594 | |
| }, | |
| { | |
| "epoch": 2.465224111282844, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.2631578947368425e-06, | |
| "loss": 0.3885, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 2.46676970633694, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.252631578947369e-06, | |
| "loss": 0.3949, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 2.4683153013910357, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 4.242105263157895e-06, | |
| "loss": 0.4085, | |
| "step": 1597 | |
| }, | |
| { | |
| "epoch": 2.469860896445131, | |
| "grad_norm": 9.875, | |
| "learning_rate": 4.2315789473684215e-06, | |
| "loss": 0.392, | |
| "step": 1598 | |
| }, | |
| { | |
| "epoch": 2.471406491499227, | |
| "grad_norm": 11.125, | |
| "learning_rate": 4.221052631578948e-06, | |
| "loss": 0.4615, | |
| "step": 1599 | |
| }, | |
| { | |
| "epoch": 2.472952086553323, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 0.3662, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.474497681607419, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.4281, | |
| "step": 1601 | |
| }, | |
| { | |
| "epoch": 2.476043276661515, | |
| "grad_norm": 10.875, | |
| "learning_rate": 4.189473684210527e-06, | |
| "loss": 0.382, | |
| "step": 1602 | |
| }, | |
| { | |
| "epoch": 2.4775888717156107, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.178947368421053e-06, | |
| "loss": 0.378, | |
| "step": 1603 | |
| }, | |
| { | |
| "epoch": 2.479134466769706, | |
| "grad_norm": 11.75, | |
| "learning_rate": 4.1684210526315794e-06, | |
| "loss": 0.436, | |
| "step": 1604 | |
| }, | |
| { | |
| "epoch": 2.480680061823802, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.157894736842106e-06, | |
| "loss": 0.3976, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 2.482225656877898, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.147368421052632e-06, | |
| "loss": 0.4102, | |
| "step": 1606 | |
| }, | |
| { | |
| "epoch": 2.483771251931994, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.136842105263158e-06, | |
| "loss": 0.3513, | |
| "step": 1607 | |
| }, | |
| { | |
| "epoch": 2.48531684698609, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 4.126315789473685e-06, | |
| "loss": 0.4326, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 2.4868624420401853, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.115789473684211e-06, | |
| "loss": 0.3978, | |
| "step": 1609 | |
| }, | |
| { | |
| "epoch": 2.488408037094281, | |
| "grad_norm": 9.625, | |
| "learning_rate": 4.105263157894737e-06, | |
| "loss": 0.3764, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.489953632148377, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 4.094736842105264e-06, | |
| "loss": 0.4472, | |
| "step": 1611 | |
| }, | |
| { | |
| "epoch": 2.491499227202473, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.08421052631579e-06, | |
| "loss": 0.3672, | |
| "step": 1612 | |
| }, | |
| { | |
| "epoch": 2.493044822256569, | |
| "grad_norm": 20.75, | |
| "learning_rate": 4.073684210526316e-06, | |
| "loss": 0.4568, | |
| "step": 1613 | |
| }, | |
| { | |
| "epoch": 2.4945904173106648, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 4.063157894736842e-06, | |
| "loss": 0.3847, | |
| "step": 1614 | |
| }, | |
| { | |
| "epoch": 2.4961360123647607, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 4.052631578947368e-06, | |
| "loss": 0.3891, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 2.497681607418856, | |
| "grad_norm": 8.75, | |
| "learning_rate": 4.042105263157895e-06, | |
| "loss": 0.3774, | |
| "step": 1616 | |
| }, | |
| { | |
| "epoch": 2.499227202472952, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.031578947368422e-06, | |
| "loss": 0.3916, | |
| "step": 1617 | |
| }, | |
| { | |
| "epoch": 2.500772797527048, | |
| "grad_norm": 10.125, | |
| "learning_rate": 4.021052631578948e-06, | |
| "loss": 0.4479, | |
| "step": 1618 | |
| }, | |
| { | |
| "epoch": 2.502318392581144, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 4.010526315789474e-06, | |
| "loss": 0.4328, | |
| "step": 1619 | |
| }, | |
| { | |
| "epoch": 2.5038639876352393, | |
| "grad_norm": 11.625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.4016, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.5054095826893352, | |
| "grad_norm": 9.375, | |
| "learning_rate": 3.989473684210526e-06, | |
| "loss": 0.3279, | |
| "step": 1621 | |
| }, | |
| { | |
| "epoch": 2.506955177743431, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 3.9789473684210525e-06, | |
| "loss": 0.418, | |
| "step": 1622 | |
| }, | |
| { | |
| "epoch": 2.508500772797527, | |
| "grad_norm": 13.375, | |
| "learning_rate": 3.968421052631579e-06, | |
| "loss": 0.4089, | |
| "step": 1623 | |
| }, | |
| { | |
| "epoch": 2.510046367851623, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 3.957894736842106e-06, | |
| "loss": 0.4321, | |
| "step": 1624 | |
| }, | |
| { | |
| "epoch": 2.511591962905719, | |
| "grad_norm": 12.875, | |
| "learning_rate": 3.947368421052632e-06, | |
| "loss": 0.4205, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 2.5131375579598147, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 3.936842105263159e-06, | |
| "loss": 0.4202, | |
| "step": 1626 | |
| }, | |
| { | |
| "epoch": 2.51468315301391, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 3.926315789473685e-06, | |
| "loss": 0.4315, | |
| "step": 1627 | |
| }, | |
| { | |
| "epoch": 2.516228748068006, | |
| "grad_norm": 10.375, | |
| "learning_rate": 3.9157894736842104e-06, | |
| "loss": 0.4529, | |
| "step": 1628 | |
| }, | |
| { | |
| "epoch": 2.517774343122102, | |
| "grad_norm": 17.125, | |
| "learning_rate": 3.905263157894737e-06, | |
| "loss": 0.4297, | |
| "step": 1629 | |
| }, | |
| { | |
| "epoch": 2.519319938176198, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 3.894736842105263e-06, | |
| "loss": 0.3866, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5208655332302934, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 3.884210526315789e-06, | |
| "loss": 0.3415, | |
| "step": 1631 | |
| }, | |
| { | |
| "epoch": 2.5224111282843893, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 3.873684210526316e-06, | |
| "loss": 0.3847, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 2.523956723338485, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 3.863157894736843e-06, | |
| "loss": 0.4469, | |
| "step": 1633 | |
| }, | |
| { | |
| "epoch": 2.525502318392581, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 3.852631578947369e-06, | |
| "loss": 0.3824, | |
| "step": 1634 | |
| }, | |
| { | |
| "epoch": 2.527047913446677, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 3.842105263157895e-06, | |
| "loss": 0.3743, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 2.528593508500773, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 3.831578947368421e-06, | |
| "loss": 0.445, | |
| "step": 1636 | |
| }, | |
| { | |
| "epoch": 2.530139103554869, | |
| "grad_norm": 10.875, | |
| "learning_rate": 3.821052631578947e-06, | |
| "loss": 0.4259, | |
| "step": 1637 | |
| }, | |
| { | |
| "epoch": 2.5316846986089647, | |
| "grad_norm": 9.5, | |
| "learning_rate": 3.810526315789474e-06, | |
| "loss": 0.4229, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 2.53323029366306, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.4429, | |
| "step": 1639 | |
| }, | |
| { | |
| "epoch": 2.534775888717156, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 3.789473684210527e-06, | |
| "loss": 0.372, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.536321483771252, | |
| "grad_norm": 10.25, | |
| "learning_rate": 3.778947368421053e-06, | |
| "loss": 0.4234, | |
| "step": 1641 | |
| }, | |
| { | |
| "epoch": 2.537867078825348, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 3.768421052631579e-06, | |
| "loss": 0.4577, | |
| "step": 1642 | |
| }, | |
| { | |
| "epoch": 2.5394126738794434, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 3.7578947368421053e-06, | |
| "loss": 0.4178, | |
| "step": 1643 | |
| }, | |
| { | |
| "epoch": 2.5409582689335393, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 3.7473684210526317e-06, | |
| "loss": 0.374, | |
| "step": 1644 | |
| }, | |
| { | |
| "epoch": 2.542503863987635, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 3.736842105263158e-06, | |
| "loss": 0.444, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 2.544049459041731, | |
| "grad_norm": 12.625, | |
| "learning_rate": 3.7263157894736848e-06, | |
| "loss": 0.4005, | |
| "step": 1646 | |
| }, | |
| { | |
| "epoch": 2.545595054095827, | |
| "grad_norm": 8.875, | |
| "learning_rate": 3.715789473684211e-06, | |
| "loss": 0.3939, | |
| "step": 1647 | |
| }, | |
| { | |
| "epoch": 2.547140649149923, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 3.7052631578947374e-06, | |
| "loss": 0.4416, | |
| "step": 1648 | |
| }, | |
| { | |
| "epoch": 2.548686244204019, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 3.6947368421052637e-06, | |
| "loss": 0.3953, | |
| "step": 1649 | |
| }, | |
| { | |
| "epoch": 2.5502318392581143, | |
| "grad_norm": 10.375, | |
| "learning_rate": 3.6842105263157896e-06, | |
| "loss": 0.3872, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.55177743431221, | |
| "grad_norm": 10.0, | |
| "learning_rate": 3.673684210526316e-06, | |
| "loss": 0.4056, | |
| "step": 1651 | |
| }, | |
| { | |
| "epoch": 2.553323029366306, | |
| "grad_norm": 10.125, | |
| "learning_rate": 3.6631578947368423e-06, | |
| "loss": 0.4306, | |
| "step": 1652 | |
| }, | |
| { | |
| "epoch": 2.554868624420402, | |
| "grad_norm": 9.0, | |
| "learning_rate": 3.6526315789473686e-06, | |
| "loss": 0.4008, | |
| "step": 1653 | |
| }, | |
| { | |
| "epoch": 2.5564142194744974, | |
| "grad_norm": 9.625, | |
| "learning_rate": 3.642105263157895e-06, | |
| "loss": 0.3627, | |
| "step": 1654 | |
| }, | |
| { | |
| "epoch": 2.5579598145285933, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 3.6315789473684217e-06, | |
| "loss": 0.4194, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 2.5595054095826892, | |
| "grad_norm": 11.75, | |
| "learning_rate": 3.621052631578948e-06, | |
| "loss": 0.4478, | |
| "step": 1656 | |
| }, | |
| { | |
| "epoch": 2.561051004636785, | |
| "grad_norm": 11.0, | |
| "learning_rate": 3.610526315789474e-06, | |
| "loss": 0.4145, | |
| "step": 1657 | |
| }, | |
| { | |
| "epoch": 2.562596599690881, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.4391, | |
| "step": 1658 | |
| }, | |
| { | |
| "epoch": 2.564142194744977, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.5894736842105266e-06, | |
| "loss": 0.4615, | |
| "step": 1659 | |
| }, | |
| { | |
| "epoch": 2.565687789799073, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 3.578947368421053e-06, | |
| "loss": 0.3776, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5672333848531683, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 3.5684210526315792e-06, | |
| "loss": 0.3906, | |
| "step": 1661 | |
| }, | |
| { | |
| "epoch": 2.5687789799072642, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 3.5578947368421056e-06, | |
| "loss": 0.4325, | |
| "step": 1662 | |
| }, | |
| { | |
| "epoch": 2.57032457496136, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 3.5473684210526323e-06, | |
| "loss": 0.3989, | |
| "step": 1663 | |
| }, | |
| { | |
| "epoch": 2.571870170015456, | |
| "grad_norm": 11.5, | |
| "learning_rate": 3.536842105263158e-06, | |
| "loss": 0.3671, | |
| "step": 1664 | |
| }, | |
| { | |
| "epoch": 2.573415765069552, | |
| "grad_norm": 9.25, | |
| "learning_rate": 3.5263157894736846e-06, | |
| "loss": 0.4, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 2.5749613601236474, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 3.515789473684211e-06, | |
| "loss": 0.3852, | |
| "step": 1666 | |
| }, | |
| { | |
| "epoch": 2.5765069551777433, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 3.505263157894737e-06, | |
| "loss": 0.4481, | |
| "step": 1667 | |
| }, | |
| { | |
| "epoch": 2.578052550231839, | |
| "grad_norm": 17.5, | |
| "learning_rate": 3.4947368421052635e-06, | |
| "loss": 0.411, | |
| "step": 1668 | |
| }, | |
| { | |
| "epoch": 2.579598145285935, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.48421052631579e-06, | |
| "loss": 0.4385, | |
| "step": 1669 | |
| }, | |
| { | |
| "epoch": 2.581143740340031, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 3.473684210526316e-06, | |
| "loss": 0.4663, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.582689335394127, | |
| "grad_norm": 27.375, | |
| "learning_rate": 3.463157894736842e-06, | |
| "loss": 0.367, | |
| "step": 1671 | |
| }, | |
| { | |
| "epoch": 2.584234930448223, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 3.4526315789473684e-06, | |
| "loss": 0.393, | |
| "step": 1672 | |
| }, | |
| { | |
| "epoch": 2.5857805255023183, | |
| "grad_norm": 12.75, | |
| "learning_rate": 3.4421052631578947e-06, | |
| "loss": 0.4084, | |
| "step": 1673 | |
| }, | |
| { | |
| "epoch": 2.587326120556414, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 3.4315789473684215e-06, | |
| "loss": 0.3922, | |
| "step": 1674 | |
| }, | |
| { | |
| "epoch": 2.58887171561051, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 3.421052631578948e-06, | |
| "loss": 0.3793, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 2.590417310664606, | |
| "grad_norm": 10.25, | |
| "learning_rate": 3.410526315789474e-06, | |
| "loss": 0.3873, | |
| "step": 1676 | |
| }, | |
| { | |
| "epoch": 2.5919629057187015, | |
| "grad_norm": 10.125, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.4841, | |
| "step": 1677 | |
| }, | |
| { | |
| "epoch": 2.5935085007727974, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 3.3894736842105264e-06, | |
| "loss": 0.4, | |
| "step": 1678 | |
| }, | |
| { | |
| "epoch": 2.5950540958268933, | |
| "grad_norm": 10.625, | |
| "learning_rate": 3.3789473684210527e-06, | |
| "loss": 0.3864, | |
| "step": 1679 | |
| }, | |
| { | |
| "epoch": 2.596599690880989, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 3.368421052631579e-06, | |
| "loss": 0.4561, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.598145285935085, | |
| "grad_norm": 10.75, | |
| "learning_rate": 3.3578947368421054e-06, | |
| "loss": 0.4422, | |
| "step": 1681 | |
| }, | |
| { | |
| "epoch": 2.599690880989181, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 3.347368421052632e-06, | |
| "loss": 0.3638, | |
| "step": 1682 | |
| }, | |
| { | |
| "epoch": 2.601236476043277, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 3.3368421052631584e-06, | |
| "loss": 0.3278, | |
| "step": 1683 | |
| }, | |
| { | |
| "epoch": 2.6027820710973724, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.3263157894736848e-06, | |
| "loss": 0.3749, | |
| "step": 1684 | |
| }, | |
| { | |
| "epoch": 2.6043276661514683, | |
| "grad_norm": 11.625, | |
| "learning_rate": 3.3157894736842107e-06, | |
| "loss": 0.4204, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 2.605873261205564, | |
| "grad_norm": 9.75, | |
| "learning_rate": 3.305263157894737e-06, | |
| "loss": 0.4236, | |
| "step": 1686 | |
| }, | |
| { | |
| "epoch": 2.60741885625966, | |
| "grad_norm": 10.875, | |
| "learning_rate": 3.2947368421052633e-06, | |
| "loss": 0.3948, | |
| "step": 1687 | |
| }, | |
| { | |
| "epoch": 2.6089644513137555, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 3.2842105263157897e-06, | |
| "loss": 0.41, | |
| "step": 1688 | |
| }, | |
| { | |
| "epoch": 2.6105100463678514, | |
| "grad_norm": 11.0, | |
| "learning_rate": 3.273684210526316e-06, | |
| "loss": 0.3952, | |
| "step": 1689 | |
| }, | |
| { | |
| "epoch": 2.6120556414219473, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 3.2631578947368423e-06, | |
| "loss": 0.402, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6136012364760433, | |
| "grad_norm": 11.25, | |
| "learning_rate": 3.252631578947369e-06, | |
| "loss": 0.3283, | |
| "step": 1691 | |
| }, | |
| { | |
| "epoch": 2.615146831530139, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 3.2421052631578945e-06, | |
| "loss": 0.3604, | |
| "step": 1692 | |
| }, | |
| { | |
| "epoch": 2.616692426584235, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 3.2315789473684213e-06, | |
| "loss": 0.4216, | |
| "step": 1693 | |
| }, | |
| { | |
| "epoch": 2.618238021638331, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 3.2210526315789476e-06, | |
| "loss": 0.4181, | |
| "step": 1694 | |
| }, | |
| { | |
| "epoch": 2.6197836166924264, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 3.210526315789474e-06, | |
| "loss": 0.4043, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 2.6213292117465223, | |
| "grad_norm": 9.0, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.3961, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 2.6228748068006182, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 3.1894736842105266e-06, | |
| "loss": 0.4476, | |
| "step": 1697 | |
| }, | |
| { | |
| "epoch": 2.624420401854714, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 3.178947368421053e-06, | |
| "loss": 0.4221, | |
| "step": 1698 | |
| }, | |
| { | |
| "epoch": 2.62596599690881, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 3.168421052631579e-06, | |
| "loss": 0.4244, | |
| "step": 1699 | |
| }, | |
| { | |
| "epoch": 2.6275115919629055, | |
| "grad_norm": 10.0, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 0.3973, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6290571870170014, | |
| "grad_norm": 11.375, | |
| "learning_rate": 3.147368421052632e-06, | |
| "loss": 0.4377, | |
| "step": 1701 | |
| }, | |
| { | |
| "epoch": 2.6306027820710973, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 3.1368421052631582e-06, | |
| "loss": 0.4146, | |
| "step": 1702 | |
| }, | |
| { | |
| "epoch": 2.6321483771251932, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 3.1263157894736846e-06, | |
| "loss": 0.4678, | |
| "step": 1703 | |
| }, | |
| { | |
| "epoch": 2.633693972179289, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 3.115789473684211e-06, | |
| "loss": 0.4078, | |
| "step": 1704 | |
| }, | |
| { | |
| "epoch": 2.635239567233385, | |
| "grad_norm": 11.125, | |
| "learning_rate": 3.1052631578947372e-06, | |
| "loss": 0.3784, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 2.636785162287481, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.094736842105263e-06, | |
| "loss": 0.4102, | |
| "step": 1706 | |
| }, | |
| { | |
| "epoch": 2.6383307573415764, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 3.0842105263157895e-06, | |
| "loss": 0.4125, | |
| "step": 1707 | |
| }, | |
| { | |
| "epoch": 2.6398763523956723, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 3.0736842105263158e-06, | |
| "loss": 0.3951, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 2.641421947449768, | |
| "grad_norm": 9.875, | |
| "learning_rate": 3.0631578947368425e-06, | |
| "loss": 0.3627, | |
| "step": 1709 | |
| }, | |
| { | |
| "epoch": 2.642967542503864, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 3.052631578947369e-06, | |
| "loss": 0.406, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6445131375579596, | |
| "grad_norm": 10.125, | |
| "learning_rate": 3.042105263157895e-06, | |
| "loss": 0.4151, | |
| "step": 1711 | |
| }, | |
| { | |
| "epoch": 2.6460587326120555, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 3.0315789473684215e-06, | |
| "loss": 0.3784, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 2.6476043276661514, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 3.0210526315789474e-06, | |
| "loss": 0.4338, | |
| "step": 1713 | |
| }, | |
| { | |
| "epoch": 2.6491499227202473, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 3.0105263157894737e-06, | |
| "loss": 0.3633, | |
| "step": 1714 | |
| }, | |
| { | |
| "epoch": 2.650695517774343, | |
| "grad_norm": 9.375, | |
| "learning_rate": 3e-06, | |
| "loss": 0.4379, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 2.652241112828439, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.9894736842105264e-06, | |
| "loss": 0.3851, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 2.653786707882535, | |
| "grad_norm": 10.5, | |
| "learning_rate": 2.9789473684210527e-06, | |
| "loss": 0.4064, | |
| "step": 1717 | |
| }, | |
| { | |
| "epoch": 2.6553323029366305, | |
| "grad_norm": 9.5, | |
| "learning_rate": 2.9684210526315795e-06, | |
| "loss": 0.3912, | |
| "step": 1718 | |
| }, | |
| { | |
| "epoch": 2.6568778979907264, | |
| "grad_norm": 10.75, | |
| "learning_rate": 2.957894736842106e-06, | |
| "loss": 0.381, | |
| "step": 1719 | |
| }, | |
| { | |
| "epoch": 2.6584234930448223, | |
| "grad_norm": 12.0, | |
| "learning_rate": 2.9473684210526317e-06, | |
| "loss": 0.4228, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.659969088098918, | |
| "grad_norm": 23.375, | |
| "learning_rate": 2.936842105263158e-06, | |
| "loss": 0.3884, | |
| "step": 1721 | |
| }, | |
| { | |
| "epoch": 2.6615146831530136, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 2.9263157894736844e-06, | |
| "loss": 0.4685, | |
| "step": 1722 | |
| }, | |
| { | |
| "epoch": 2.6630602782071096, | |
| "grad_norm": 10.125, | |
| "learning_rate": 2.9157894736842107e-06, | |
| "loss": 0.3619, | |
| "step": 1723 | |
| }, | |
| { | |
| "epoch": 2.6646058732612055, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 2.905263157894737e-06, | |
| "loss": 0.3738, | |
| "step": 1724 | |
| }, | |
| { | |
| "epoch": 2.6661514683153014, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 2.8947368421052634e-06, | |
| "loss": 0.3707, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 2.6676970633693973, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 2.88421052631579e-06, | |
| "loss": 0.3928, | |
| "step": 1726 | |
| }, | |
| { | |
| "epoch": 2.669242658423493, | |
| "grad_norm": 13.0, | |
| "learning_rate": 2.8736842105263164e-06, | |
| "loss": 0.3825, | |
| "step": 1727 | |
| }, | |
| { | |
| "epoch": 2.670788253477589, | |
| "grad_norm": 10.375, | |
| "learning_rate": 2.8631578947368423e-06, | |
| "loss": 0.3854, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 2.6723338485316845, | |
| "grad_norm": 10.75, | |
| "learning_rate": 2.8526315789473687e-06, | |
| "loss": 0.3673, | |
| "step": 1729 | |
| }, | |
| { | |
| "epoch": 2.6738794435857804, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 2.842105263157895e-06, | |
| "loss": 0.4554, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6754250386398764, | |
| "grad_norm": 9.5, | |
| "learning_rate": 2.8315789473684213e-06, | |
| "loss": 0.3797, | |
| "step": 1731 | |
| }, | |
| { | |
| "epoch": 2.6769706336939723, | |
| "grad_norm": 9.375, | |
| "learning_rate": 2.8210526315789476e-06, | |
| "loss": 0.4412, | |
| "step": 1732 | |
| }, | |
| { | |
| "epoch": 2.678516228748068, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 2.810526315789474e-06, | |
| "loss": 0.4124, | |
| "step": 1733 | |
| }, | |
| { | |
| "epoch": 2.6800618238021636, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.3944, | |
| "step": 1734 | |
| }, | |
| { | |
| "epoch": 2.6816074188562595, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.789473684210526e-06, | |
| "loss": 0.3976, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 2.6831530139103554, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 2.7789473684210525e-06, | |
| "loss": 0.3777, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 2.6846986089644513, | |
| "grad_norm": 12.5, | |
| "learning_rate": 2.7684210526315793e-06, | |
| "loss": 0.4489, | |
| "step": 1737 | |
| }, | |
| { | |
| "epoch": 2.6862442040185472, | |
| "grad_norm": 14.625, | |
| "learning_rate": 2.7578947368421056e-06, | |
| "loss": 0.4448, | |
| "step": 1738 | |
| }, | |
| { | |
| "epoch": 2.687789799072643, | |
| "grad_norm": 10.5, | |
| "learning_rate": 2.747368421052632e-06, | |
| "loss": 0.3733, | |
| "step": 1739 | |
| }, | |
| { | |
| "epoch": 2.689335394126739, | |
| "grad_norm": 9.125, | |
| "learning_rate": 2.7368421052631583e-06, | |
| "loss": 0.3982, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.6908809891808345, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.7263157894736846e-06, | |
| "loss": 0.4048, | |
| "step": 1741 | |
| }, | |
| { | |
| "epoch": 2.6924265842349304, | |
| "grad_norm": 9.75, | |
| "learning_rate": 2.7157894736842105e-06, | |
| "loss": 0.408, | |
| "step": 1742 | |
| }, | |
| { | |
| "epoch": 2.6939721792890263, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.705263157894737e-06, | |
| "loss": 0.4068, | |
| "step": 1743 | |
| }, | |
| { | |
| "epoch": 2.6955177743431222, | |
| "grad_norm": 10.125, | |
| "learning_rate": 2.694736842105263e-06, | |
| "loss": 0.4276, | |
| "step": 1744 | |
| }, | |
| { | |
| "epoch": 2.6970633693972177, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 2.68421052631579e-06, | |
| "loss": 0.4537, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 2.6986089644513136, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 2.6736842105263162e-06, | |
| "loss": 0.4682, | |
| "step": 1746 | |
| }, | |
| { | |
| "epoch": 2.7001545595054095, | |
| "grad_norm": 10.125, | |
| "learning_rate": 2.6631578947368426e-06, | |
| "loss": 0.3858, | |
| "step": 1747 | |
| }, | |
| { | |
| "epoch": 2.7017001545595054, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.652631578947369e-06, | |
| "loss": 0.3793, | |
| "step": 1748 | |
| }, | |
| { | |
| "epoch": 2.7032457496136013, | |
| "grad_norm": 10.75, | |
| "learning_rate": 2.6421052631578948e-06, | |
| "loss": 0.395, | |
| "step": 1749 | |
| }, | |
| { | |
| "epoch": 2.704791344667697, | |
| "grad_norm": 10.75, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 0.3849, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.706336939721793, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.6210526315789474e-06, | |
| "loss": 0.4047, | |
| "step": 1751 | |
| }, | |
| { | |
| "epoch": 2.7078825347758886, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 2.6105263157894738e-06, | |
| "loss": 0.4146, | |
| "step": 1752 | |
| }, | |
| { | |
| "epoch": 2.7094281298299845, | |
| "grad_norm": 11.375, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.4149, | |
| "step": 1753 | |
| }, | |
| { | |
| "epoch": 2.7109737248840804, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 2.589473684210527e-06, | |
| "loss": 0.3694, | |
| "step": 1754 | |
| }, | |
| { | |
| "epoch": 2.7125193199381763, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.578947368421053e-06, | |
| "loss": 0.4454, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 2.7140649149922718, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 2.568421052631579e-06, | |
| "loss": 0.3473, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 2.7156105100463677, | |
| "grad_norm": 9.875, | |
| "learning_rate": 2.5578947368421054e-06, | |
| "loss": 0.3919, | |
| "step": 1757 | |
| }, | |
| { | |
| "epoch": 2.7171561051004636, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.5473684210526317e-06, | |
| "loss": 0.3689, | |
| "step": 1758 | |
| }, | |
| { | |
| "epoch": 2.7187017001545595, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 2.536842105263158e-06, | |
| "loss": 0.3767, | |
| "step": 1759 | |
| }, | |
| { | |
| "epoch": 2.7202472952086554, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 2.5263157894736844e-06, | |
| "loss": 0.4031, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.7217928902627513, | |
| "grad_norm": 10.75, | |
| "learning_rate": 2.5157894736842107e-06, | |
| "loss": 0.4077, | |
| "step": 1761 | |
| }, | |
| { | |
| "epoch": 2.723338485316847, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 2.5052631578947375e-06, | |
| "loss": 0.4016, | |
| "step": 1762 | |
| }, | |
| { | |
| "epoch": 2.7248840803709427, | |
| "grad_norm": 9.625, | |
| "learning_rate": 2.4947368421052634e-06, | |
| "loss": 0.3557, | |
| "step": 1763 | |
| }, | |
| { | |
| "epoch": 2.7264296754250386, | |
| "grad_norm": 10.625, | |
| "learning_rate": 2.4842105263157897e-06, | |
| "loss": 0.3676, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 2.7279752704791345, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.473684210526316e-06, | |
| "loss": 0.4012, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 2.7295208655332304, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 2.4631578947368424e-06, | |
| "loss": 0.3387, | |
| "step": 1766 | |
| }, | |
| { | |
| "epoch": 2.7310664605873263, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.4526315789473687e-06, | |
| "loss": 0.3777, | |
| "step": 1767 | |
| }, | |
| { | |
| "epoch": 2.7326120556414217, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.442105263157895e-06, | |
| "loss": 0.4259, | |
| "step": 1768 | |
| }, | |
| { | |
| "epoch": 2.7341576506955176, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.4315789473684213e-06, | |
| "loss": 0.3763, | |
| "step": 1769 | |
| }, | |
| { | |
| "epoch": 2.7357032457496135, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 2.4210526315789477e-06, | |
| "loss": 0.4146, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7372488408037094, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 2.410526315789474e-06, | |
| "loss": 0.4534, | |
| "step": 1771 | |
| }, | |
| { | |
| "epoch": 2.7387944358578054, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.3773, | |
| "step": 1772 | |
| }, | |
| { | |
| "epoch": 2.7403400309119013, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 2.3894736842105266e-06, | |
| "loss": 0.3715, | |
| "step": 1773 | |
| }, | |
| { | |
| "epoch": 2.741885625965997, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 2.378947368421053e-06, | |
| "loss": 0.4526, | |
| "step": 1774 | |
| }, | |
| { | |
| "epoch": 2.7434312210200926, | |
| "grad_norm": 8.75, | |
| "learning_rate": 2.368421052631579e-06, | |
| "loss": 0.3926, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 2.7449768160741885, | |
| "grad_norm": 10.75, | |
| "learning_rate": 2.357894736842105e-06, | |
| "loss": 0.434, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 2.7465224111282844, | |
| "grad_norm": 20.625, | |
| "learning_rate": 2.347368421052632e-06, | |
| "loss": 0.4384, | |
| "step": 1777 | |
| }, | |
| { | |
| "epoch": 2.7480680061823803, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.3368421052631583e-06, | |
| "loss": 0.372, | |
| "step": 1778 | |
| }, | |
| { | |
| "epoch": 2.749613601236476, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 2.326315789473684e-06, | |
| "loss": 0.4376, | |
| "step": 1779 | |
| }, | |
| { | |
| "epoch": 2.7511591962905717, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 2.3157894736842105e-06, | |
| "loss": 0.3777, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.7527047913446676, | |
| "grad_norm": 11.375, | |
| "learning_rate": 2.3052631578947373e-06, | |
| "loss": 0.4423, | |
| "step": 1781 | |
| }, | |
| { | |
| "epoch": 2.7542503863987635, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 2.294736842105263e-06, | |
| "loss": 0.3894, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 2.7557959814528594, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.2842105263157895e-06, | |
| "loss": 0.3711, | |
| "step": 1783 | |
| }, | |
| { | |
| "epoch": 2.7573415765069553, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.273684210526316e-06, | |
| "loss": 0.3663, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.7588871715610512, | |
| "grad_norm": 10.125, | |
| "learning_rate": 2.2631578947368426e-06, | |
| "loss": 0.3994, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 2.7604327666151467, | |
| "grad_norm": 9.125, | |
| "learning_rate": 2.2526315789473685e-06, | |
| "loss": 0.3846, | |
| "step": 1786 | |
| }, | |
| { | |
| "epoch": 2.7619783616692426, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.242105263157895e-06, | |
| "loss": 0.3311, | |
| "step": 1787 | |
| }, | |
| { | |
| "epoch": 2.7635239567233385, | |
| "grad_norm": 12.125, | |
| "learning_rate": 2.231578947368421e-06, | |
| "loss": 0.4519, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 2.7650695517774344, | |
| "grad_norm": 10.875, | |
| "learning_rate": 2.221052631578948e-06, | |
| "loss": 0.4599, | |
| "step": 1789 | |
| }, | |
| { | |
| "epoch": 2.76661514683153, | |
| "grad_norm": 31.625, | |
| "learning_rate": 2.2105263157894738e-06, | |
| "loss": 0.3943, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.7681607418856258, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.4059, | |
| "step": 1791 | |
| }, | |
| { | |
| "epoch": 2.7697063369397217, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 2.1894736842105264e-06, | |
| "loss": 0.3567, | |
| "step": 1792 | |
| }, | |
| { | |
| "epoch": 2.7712519319938176, | |
| "grad_norm": 10.125, | |
| "learning_rate": 2.1789473684210528e-06, | |
| "loss": 0.3779, | |
| "step": 1793 | |
| }, | |
| { | |
| "epoch": 2.7727975270479135, | |
| "grad_norm": 9.375, | |
| "learning_rate": 2.168421052631579e-06, | |
| "loss": 0.3605, | |
| "step": 1794 | |
| }, | |
| { | |
| "epoch": 2.7743431221020094, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.1578947368421054e-06, | |
| "loss": 0.3392, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 2.7758887171561053, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 2.1473684210526317e-06, | |
| "loss": 0.3849, | |
| "step": 1796 | |
| }, | |
| { | |
| "epoch": 2.7774343122102008, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.136842105263158e-06, | |
| "loss": 0.4181, | |
| "step": 1797 | |
| }, | |
| { | |
| "epoch": 2.7789799072642967, | |
| "grad_norm": 12.25, | |
| "learning_rate": 2.1263157894736844e-06, | |
| "loss": 0.4537, | |
| "step": 1798 | |
| }, | |
| { | |
| "epoch": 2.7805255023183926, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.1157894736842107e-06, | |
| "loss": 0.3912, | |
| "step": 1799 | |
| }, | |
| { | |
| "epoch": 2.7820710973724885, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 0.4458, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.7836166924265844, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.0947368421052634e-06, | |
| "loss": 0.3801, | |
| "step": 1801 | |
| }, | |
| { | |
| "epoch": 2.78516228748068, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.0842105263157897e-06, | |
| "loss": 0.3761, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 2.7867078825347757, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 2.073684210526316e-06, | |
| "loss": 0.4046, | |
| "step": 1803 | |
| }, | |
| { | |
| "epoch": 2.7882534775888717, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 2.0631578947368424e-06, | |
| "loss": 0.3733, | |
| "step": 1804 | |
| }, | |
| { | |
| "epoch": 2.7897990726429676, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 2.0526315789473687e-06, | |
| "loss": 0.4184, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 2.7913446676970635, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 2.042105263157895e-06, | |
| "loss": 0.4501, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 2.7928902627511594, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 2.031578947368421e-06, | |
| "loss": 0.4095, | |
| "step": 1807 | |
| }, | |
| { | |
| "epoch": 2.7944358578052553, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 2.0210526315789477e-06, | |
| "loss": 0.4088, | |
| "step": 1808 | |
| }, | |
| { | |
| "epoch": 2.7959814528593507, | |
| "grad_norm": 11.875, | |
| "learning_rate": 2.010526315789474e-06, | |
| "loss": 0.4242, | |
| "step": 1809 | |
| }, | |
| { | |
| "epoch": 2.7975270479134466, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.3917, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.7990726429675425, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 1.9894736842105262e-06, | |
| "loss": 0.4334, | |
| "step": 1811 | |
| }, | |
| { | |
| "epoch": 2.8006182380216385, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.978947368421053e-06, | |
| "loss": 0.483, | |
| "step": 1812 | |
| }, | |
| { | |
| "epoch": 2.802163833075734, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.9684210526315793e-06, | |
| "loss": 0.4119, | |
| "step": 1813 | |
| }, | |
| { | |
| "epoch": 2.80370942812983, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.9578947368421052e-06, | |
| "loss": 0.4372, | |
| "step": 1814 | |
| }, | |
| { | |
| "epoch": 2.8052550231839257, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.9473684210526315e-06, | |
| "loss": 0.3909, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 2.8068006182380216, | |
| "grad_norm": 16.375, | |
| "learning_rate": 1.936842105263158e-06, | |
| "loss": 0.4327, | |
| "step": 1816 | |
| }, | |
| { | |
| "epoch": 2.8083462132921175, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.9263157894736846e-06, | |
| "loss": 0.4093, | |
| "step": 1817 | |
| }, | |
| { | |
| "epoch": 2.8098918083462134, | |
| "grad_norm": 13.625, | |
| "learning_rate": 1.9157894736842105e-06, | |
| "loss": 0.4482, | |
| "step": 1818 | |
| }, | |
| { | |
| "epoch": 2.8114374034003093, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.905263157894737e-06, | |
| "loss": 0.3899, | |
| "step": 1819 | |
| }, | |
| { | |
| "epoch": 2.812982998454405, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.8947368421052634e-06, | |
| "loss": 0.4006, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8145285935085007, | |
| "grad_norm": 11.75, | |
| "learning_rate": 1.8842105263157895e-06, | |
| "loss": 0.3468, | |
| "step": 1821 | |
| }, | |
| { | |
| "epoch": 2.8160741885625966, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.8736842105263158e-06, | |
| "loss": 0.4039, | |
| "step": 1822 | |
| }, | |
| { | |
| "epoch": 2.8176197836166925, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.8631578947368424e-06, | |
| "loss": 0.3988, | |
| "step": 1823 | |
| }, | |
| { | |
| "epoch": 2.819165378670788, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.8526315789473687e-06, | |
| "loss": 0.4123, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 2.820710973724884, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.8421052631578948e-06, | |
| "loss": 0.4064, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 2.82225656877898, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.8315789473684211e-06, | |
| "loss": 0.4234, | |
| "step": 1826 | |
| }, | |
| { | |
| "epoch": 2.8238021638330757, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.8210526315789475e-06, | |
| "loss": 0.3731, | |
| "step": 1827 | |
| }, | |
| { | |
| "epoch": 2.8253477588871716, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.810526315789474e-06, | |
| "loss": 0.3949, | |
| "step": 1828 | |
| }, | |
| { | |
| "epoch": 2.8268933539412675, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.3817, | |
| "step": 1829 | |
| }, | |
| { | |
| "epoch": 2.8284389489953634, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 1.7894736842105265e-06, | |
| "loss": 0.4308, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.8299845440494593, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 1.7789473684210528e-06, | |
| "loss": 0.3637, | |
| "step": 1831 | |
| }, | |
| { | |
| "epoch": 2.8315301391035548, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.768421052631579e-06, | |
| "loss": 0.3711, | |
| "step": 1832 | |
| }, | |
| { | |
| "epoch": 2.8330757341576507, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.7578947368421054e-06, | |
| "loss": 0.4294, | |
| "step": 1833 | |
| }, | |
| { | |
| "epoch": 2.8346213292117466, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.7473684210526318e-06, | |
| "loss": 0.3973, | |
| "step": 1834 | |
| }, | |
| { | |
| "epoch": 2.8361669242658425, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.736842105263158e-06, | |
| "loss": 0.4143, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 2.837712519319938, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 1.7263157894736842e-06, | |
| "loss": 0.4742, | |
| "step": 1836 | |
| }, | |
| { | |
| "epoch": 2.839258114374034, | |
| "grad_norm": 17.0, | |
| "learning_rate": 1.7157894736842107e-06, | |
| "loss": 0.4588, | |
| "step": 1837 | |
| }, | |
| { | |
| "epoch": 2.8408037094281298, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.705263157894737e-06, | |
| "loss": 0.4121, | |
| "step": 1838 | |
| }, | |
| { | |
| "epoch": 2.8423493044822257, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.6947368421052632e-06, | |
| "loss": 0.4282, | |
| "step": 1839 | |
| }, | |
| { | |
| "epoch": 2.8438948995363216, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.6842105263157895e-06, | |
| "loss": 0.4076, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.8454404945904175, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.673684210526316e-06, | |
| "loss": 0.3933, | |
| "step": 1841 | |
| }, | |
| { | |
| "epoch": 2.8469860896445134, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.6631578947368424e-06, | |
| "loss": 0.4362, | |
| "step": 1842 | |
| }, | |
| { | |
| "epoch": 2.848531684698609, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.6526315789473685e-06, | |
| "loss": 0.4491, | |
| "step": 1843 | |
| }, | |
| { | |
| "epoch": 2.8500772797527048, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.6421052631578948e-06, | |
| "loss": 0.3365, | |
| "step": 1844 | |
| }, | |
| { | |
| "epoch": 2.8516228748068007, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.6315789473684212e-06, | |
| "loss": 0.392, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 2.8531684698608966, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 1.6210526315789473e-06, | |
| "loss": 0.4085, | |
| "step": 1846 | |
| }, | |
| { | |
| "epoch": 2.854714064914992, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.6105263157894738e-06, | |
| "loss": 0.3727, | |
| "step": 1847 | |
| }, | |
| { | |
| "epoch": 2.856259659969088, | |
| "grad_norm": 48.25, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.4389, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 2.857805255023184, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.5894736842105265e-06, | |
| "loss": 0.4017, | |
| "step": 1849 | |
| }, | |
| { | |
| "epoch": 2.8593508500772797, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.5789473684210526e-06, | |
| "loss": 0.3863, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.8608964451313756, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.5684210526315791e-06, | |
| "loss": 0.3702, | |
| "step": 1851 | |
| }, | |
| { | |
| "epoch": 2.8624420401854715, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.5578947368421054e-06, | |
| "loss": 0.4564, | |
| "step": 1852 | |
| }, | |
| { | |
| "epoch": 2.8639876352395675, | |
| "grad_norm": 13.5, | |
| "learning_rate": 1.5473684210526316e-06, | |
| "loss": 0.3896, | |
| "step": 1853 | |
| }, | |
| { | |
| "epoch": 2.865533230293663, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.5368421052631579e-06, | |
| "loss": 0.4136, | |
| "step": 1854 | |
| }, | |
| { | |
| "epoch": 2.867078825347759, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 1.5263157894736844e-06, | |
| "loss": 0.4244, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 2.8686244204018547, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.5157894736842108e-06, | |
| "loss": 0.4011, | |
| "step": 1856 | |
| }, | |
| { | |
| "epoch": 2.8701700154559506, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.5052631578947369e-06, | |
| "loss": 0.4049, | |
| "step": 1857 | |
| }, | |
| { | |
| "epoch": 2.871715610510046, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.4947368421052632e-06, | |
| "loss": 0.399, | |
| "step": 1858 | |
| }, | |
| { | |
| "epoch": 2.873261205564142, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 1.4842105263157897e-06, | |
| "loss": 0.4574, | |
| "step": 1859 | |
| }, | |
| { | |
| "epoch": 2.874806800618238, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.4736842105263159e-06, | |
| "loss": 0.3981, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.876352395672334, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.4631578947368422e-06, | |
| "loss": 0.458, | |
| "step": 1861 | |
| }, | |
| { | |
| "epoch": 2.8778979907264297, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.4526315789473685e-06, | |
| "loss": 0.4053, | |
| "step": 1862 | |
| }, | |
| { | |
| "epoch": 2.8794435857805256, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.442105263157895e-06, | |
| "loss": 0.4028, | |
| "step": 1863 | |
| }, | |
| { | |
| "epoch": 2.8809891808346215, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.4315789473684212e-06, | |
| "loss": 0.453, | |
| "step": 1864 | |
| }, | |
| { | |
| "epoch": 2.8825347758887174, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.4210526315789475e-06, | |
| "loss": 0.4118, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 2.884080370942813, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.4105263157894738e-06, | |
| "loss": 0.3685, | |
| "step": 1866 | |
| }, | |
| { | |
| "epoch": 2.885625965996909, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.3809, | |
| "step": 1867 | |
| }, | |
| { | |
| "epoch": 2.8871715610510047, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.3894736842105263e-06, | |
| "loss": 0.4276, | |
| "step": 1868 | |
| }, | |
| { | |
| "epoch": 2.8887171561051006, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.3789473684210528e-06, | |
| "loss": 0.4094, | |
| "step": 1869 | |
| }, | |
| { | |
| "epoch": 2.890262751159196, | |
| "grad_norm": 14.625, | |
| "learning_rate": 1.3684210526315791e-06, | |
| "loss": 0.4412, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.891808346213292, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.3578947368421052e-06, | |
| "loss": 0.3844, | |
| "step": 1871 | |
| }, | |
| { | |
| "epoch": 2.893353941267388, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.3473684210526316e-06, | |
| "loss": 0.4251, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 2.894899536321484, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.3368421052631581e-06, | |
| "loss": 0.371, | |
| "step": 1873 | |
| }, | |
| { | |
| "epoch": 2.8964451313755797, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.3263157894736844e-06, | |
| "loss": 0.4371, | |
| "step": 1874 | |
| }, | |
| { | |
| "epoch": 2.8979907264296756, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.3157894736842106e-06, | |
| "loss": 0.427, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.8995363214837715, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.3052631578947369e-06, | |
| "loss": 0.4068, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 2.901081916537867, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.2947368421052634e-06, | |
| "loss": 0.3963, | |
| "step": 1877 | |
| }, | |
| { | |
| "epoch": 2.902627511591963, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.2842105263157895e-06, | |
| "loss": 0.3563, | |
| "step": 1878 | |
| }, | |
| { | |
| "epoch": 2.9041731066460588, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.2736842105263159e-06, | |
| "loss": 0.4034, | |
| "step": 1879 | |
| }, | |
| { | |
| "epoch": 2.9057187017001547, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.2631578947368422e-06, | |
| "loss": 0.3989, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.90726429675425, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.2526315789473687e-06, | |
| "loss": 0.3692, | |
| "step": 1881 | |
| }, | |
| { | |
| "epoch": 2.908809891808346, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2421052631578948e-06, | |
| "loss": 0.4374, | |
| "step": 1882 | |
| }, | |
| { | |
| "epoch": 2.910355486862442, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.2315789473684212e-06, | |
| "loss": 0.4132, | |
| "step": 1883 | |
| }, | |
| { | |
| "epoch": 2.911901081916538, | |
| "grad_norm": 13.25, | |
| "learning_rate": 1.2210526315789475e-06, | |
| "loss": 0.3918, | |
| "step": 1884 | |
| }, | |
| { | |
| "epoch": 2.9134466769706338, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.2105263157894738e-06, | |
| "loss": 0.385, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 2.9149922720247297, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.3525, | |
| "step": 1886 | |
| }, | |
| { | |
| "epoch": 2.9165378670788256, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.1894736842105265e-06, | |
| "loss": 0.3566, | |
| "step": 1887 | |
| }, | |
| { | |
| "epoch": 2.918083462132921, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.1789473684210526e-06, | |
| "loss": 0.4376, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 2.919629057187017, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.1684210526315791e-06, | |
| "loss": 0.4202, | |
| "step": 1889 | |
| }, | |
| { | |
| "epoch": 2.921174652241113, | |
| "grad_norm": 8.75, | |
| "learning_rate": 1.1578947368421053e-06, | |
| "loss": 0.3859, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9227202472952087, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.1473684210526316e-06, | |
| "loss": 0.4488, | |
| "step": 1891 | |
| }, | |
| { | |
| "epoch": 2.9242658423493046, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.136842105263158e-06, | |
| "loss": 0.4885, | |
| "step": 1892 | |
| }, | |
| { | |
| "epoch": 2.9258114374034, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.1263157894736842e-06, | |
| "loss": 0.4429, | |
| "step": 1893 | |
| }, | |
| { | |
| "epoch": 2.927357032457496, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.1157894736842106e-06, | |
| "loss": 0.4645, | |
| "step": 1894 | |
| }, | |
| { | |
| "epoch": 2.928902627511592, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.1052631578947369e-06, | |
| "loss": 0.3797, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 2.930448222565688, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 1.0947368421052632e-06, | |
| "loss": 0.4121, | |
| "step": 1896 | |
| }, | |
| { | |
| "epoch": 2.9319938176197837, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.0842105263157895e-06, | |
| "loss": 0.4085, | |
| "step": 1897 | |
| }, | |
| { | |
| "epoch": 2.9335394126738796, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.0736842105263159e-06, | |
| "loss": 0.3847, | |
| "step": 1898 | |
| }, | |
| { | |
| "epoch": 2.9350850077279755, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.0631578947368422e-06, | |
| "loss": 0.4126, | |
| "step": 1899 | |
| }, | |
| { | |
| "epoch": 2.936630602782071, | |
| "grad_norm": 10.75, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 0.4209, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.938176197836167, | |
| "grad_norm": 12.125, | |
| "learning_rate": 1.0421052631578949e-06, | |
| "loss": 0.4829, | |
| "step": 1901 | |
| }, | |
| { | |
| "epoch": 2.939721792890263, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.0315789473684212e-06, | |
| "loss": 0.4663, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 2.9412673879443587, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 1.0210526315789475e-06, | |
| "loss": 0.4614, | |
| "step": 1903 | |
| }, | |
| { | |
| "epoch": 2.942812982998454, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.0105263157894738e-06, | |
| "loss": 0.4151, | |
| "step": 1904 | |
| }, | |
| { | |
| "epoch": 2.94435857805255, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.4042, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 2.945904173106646, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 9.894736842105265e-07, | |
| "loss": 0.4206, | |
| "step": 1906 | |
| }, | |
| { | |
| "epoch": 2.947449768160742, | |
| "grad_norm": 14.8125, | |
| "learning_rate": 9.789473684210526e-07, | |
| "loss": 0.4037, | |
| "step": 1907 | |
| }, | |
| { | |
| "epoch": 2.948995363214838, | |
| "grad_norm": 9.75, | |
| "learning_rate": 9.68421052631579e-07, | |
| "loss": 0.3797, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 2.9505409582689337, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.578947368421053e-07, | |
| "loss": 0.3993, | |
| "step": 1909 | |
| }, | |
| { | |
| "epoch": 2.9520865533230296, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.473684210526317e-07, | |
| "loss": 0.4062, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.953632148377125, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.368421052631579e-07, | |
| "loss": 0.3637, | |
| "step": 1911 | |
| }, | |
| { | |
| "epoch": 2.955177743431221, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.263157894736844e-07, | |
| "loss": 0.3892, | |
| "step": 1912 | |
| }, | |
| { | |
| "epoch": 2.956723338485317, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.157894736842106e-07, | |
| "loss": 0.4085, | |
| "step": 1913 | |
| }, | |
| { | |
| "epoch": 2.958268933539413, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.05263157894737e-07, | |
| "loss": 0.4408, | |
| "step": 1914 | |
| }, | |
| { | |
| "epoch": 2.9598145285935082, | |
| "grad_norm": 10.5, | |
| "learning_rate": 8.947368421052632e-07, | |
| "loss": 0.3937, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 2.961360123647604, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 8.842105263157895e-07, | |
| "loss": 0.43, | |
| "step": 1916 | |
| }, | |
| { | |
| "epoch": 2.9629057187017, | |
| "grad_norm": 9.75, | |
| "learning_rate": 8.736842105263159e-07, | |
| "loss": 0.3885, | |
| "step": 1917 | |
| }, | |
| { | |
| "epoch": 2.964451313755796, | |
| "grad_norm": 13.875, | |
| "learning_rate": 8.631578947368421e-07, | |
| "loss": 0.3249, | |
| "step": 1918 | |
| }, | |
| { | |
| "epoch": 2.965996908809892, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 8.526315789473685e-07, | |
| "loss": 0.4174, | |
| "step": 1919 | |
| }, | |
| { | |
| "epoch": 2.9675425038639878, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 8.421052631578948e-07, | |
| "loss": 0.3626, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.9690880989180837, | |
| "grad_norm": 9.625, | |
| "learning_rate": 8.315789473684212e-07, | |
| "loss": 0.3434, | |
| "step": 1921 | |
| }, | |
| { | |
| "epoch": 2.970633693972179, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 8.210526315789474e-07, | |
| "loss": 0.3301, | |
| "step": 1922 | |
| }, | |
| { | |
| "epoch": 2.972179289026275, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 8.105263157894736e-07, | |
| "loss": 0.3664, | |
| "step": 1923 | |
| }, | |
| { | |
| "epoch": 2.973724884080371, | |
| "grad_norm": 11.625, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.3901, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 2.975270479134467, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.894736842105263e-07, | |
| "loss": 0.3904, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.9768160741885628, | |
| "grad_norm": 9.875, | |
| "learning_rate": 7.789473684210527e-07, | |
| "loss": 0.4489, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 2.978361669242658, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 7.684210526315789e-07, | |
| "loss": 0.3721, | |
| "step": 1927 | |
| }, | |
| { | |
| "epoch": 2.979907264296754, | |
| "grad_norm": 17.875, | |
| "learning_rate": 7.578947368421054e-07, | |
| "loss": 0.3797, | |
| "step": 1928 | |
| }, | |
| { | |
| "epoch": 2.98145285935085, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 7.473684210526316e-07, | |
| "loss": 0.4435, | |
| "step": 1929 | |
| }, | |
| { | |
| "epoch": 2.982998454404946, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 7.368421052631579e-07, | |
| "loss": 0.3561, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.984544049459042, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 7.263157894736843e-07, | |
| "loss": 0.3817, | |
| "step": 1931 | |
| }, | |
| { | |
| "epoch": 2.9860896445131377, | |
| "grad_norm": 9.875, | |
| "learning_rate": 7.157894736842106e-07, | |
| "loss": 0.4161, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 2.9876352395672336, | |
| "grad_norm": 10.5, | |
| "learning_rate": 7.052631578947369e-07, | |
| "loss": 0.3882, | |
| "step": 1933 | |
| }, | |
| { | |
| "epoch": 2.989180834621329, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 6.947368421052631e-07, | |
| "loss": 0.3813, | |
| "step": 1934 | |
| }, | |
| { | |
| "epoch": 2.990726429675425, | |
| "grad_norm": 10.75, | |
| "learning_rate": 6.842105263157896e-07, | |
| "loss": 0.4273, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 2.992272024729521, | |
| "grad_norm": 11.25, | |
| "learning_rate": 6.736842105263158e-07, | |
| "loss": 0.4464, | |
| "step": 1936 | |
| }, | |
| { | |
| "epoch": 2.993817619783617, | |
| "grad_norm": 9.75, | |
| "learning_rate": 6.631578947368422e-07, | |
| "loss": 0.4175, | |
| "step": 1937 | |
| }, | |
| { | |
| "epoch": 2.9953632148377123, | |
| "grad_norm": 10.625, | |
| "learning_rate": 6.526315789473684e-07, | |
| "loss": 0.4584, | |
| "step": 1938 | |
| }, | |
| { | |
| "epoch": 2.996908809891808, | |
| "grad_norm": 9.875, | |
| "learning_rate": 6.421052631578948e-07, | |
| "loss": 0.4013, | |
| "step": 1939 | |
| }, | |
| { | |
| "epoch": 2.998454404945904, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 6.315789473684211e-07, | |
| "loss": 0.4026, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 8.625, | |
| "learning_rate": 6.210526315789474e-07, | |
| "loss": 0.3594, | |
| "step": 1941 | |
| }, | |
| { | |
| "epoch": 3.001545595054096, | |
| "grad_norm": 9.25, | |
| "learning_rate": 6.105263157894738e-07, | |
| "loss": 0.3907, | |
| "step": 1942 | |
| }, | |
| { | |
| "epoch": 3.003091190108192, | |
| "grad_norm": 8.5, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.3545, | |
| "step": 1943 | |
| }, | |
| { | |
| "epoch": 3.0046367851622873, | |
| "grad_norm": 10.375, | |
| "learning_rate": 5.894736842105263e-07, | |
| "loss": 0.3784, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 3.006182380216383, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 5.789473684210526e-07, | |
| "loss": 0.3808, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 3.007727975270479, | |
| "grad_norm": 14.125, | |
| "learning_rate": 5.68421052631579e-07, | |
| "loss": 0.4312, | |
| "step": 1946 | |
| }, | |
| { | |
| "epoch": 3.009273570324575, | |
| "grad_norm": 10.5, | |
| "learning_rate": 5.578947368421053e-07, | |
| "loss": 0.3974, | |
| "step": 1947 | |
| }, | |
| { | |
| "epoch": 3.010819165378671, | |
| "grad_norm": 28.125, | |
| "learning_rate": 5.473684210526316e-07, | |
| "loss": 0.4432, | |
| "step": 1948 | |
| }, | |
| { | |
| "epoch": 3.012364760432767, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 5.368421052631579e-07, | |
| "loss": 0.3728, | |
| "step": 1949 | |
| }, | |
| { | |
| "epoch": 3.0139103554868623, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 5.263157894736843e-07, | |
| "loss": 0.3761, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.015455950540958, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 5.157894736842106e-07, | |
| "loss": 0.3817, | |
| "step": 1951 | |
| }, | |
| { | |
| "epoch": 3.017001545595054, | |
| "grad_norm": 8.875, | |
| "learning_rate": 5.052631578947369e-07, | |
| "loss": 0.3743, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 3.01854714064915, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 4.947368421052632e-07, | |
| "loss": 0.4283, | |
| "step": 1953 | |
| }, | |
| { | |
| "epoch": 3.020092735703246, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 4.842105263157895e-07, | |
| "loss": 0.4342, | |
| "step": 1954 | |
| }, | |
| { | |
| "epoch": 3.021638330757342, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.7368421052631585e-07, | |
| "loss": 0.3495, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 3.0231839258114372, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 4.631578947368422e-07, | |
| "loss": 0.4606, | |
| "step": 1956 | |
| }, | |
| { | |
| "epoch": 3.024729520865533, | |
| "grad_norm": 11.75, | |
| "learning_rate": 4.526315789473685e-07, | |
| "loss": 0.3993, | |
| "step": 1957 | |
| }, | |
| { | |
| "epoch": 3.026275115919629, | |
| "grad_norm": 11.375, | |
| "learning_rate": 4.421052631578947e-07, | |
| "loss": 0.4741, | |
| "step": 1958 | |
| }, | |
| { | |
| "epoch": 3.027820710973725, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.3157894736842105e-07, | |
| "loss": 0.381, | |
| "step": 1959 | |
| }, | |
| { | |
| "epoch": 3.029366306027821, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.210526315789474e-07, | |
| "loss": 0.3736, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0309119010819163, | |
| "grad_norm": 11.75, | |
| "learning_rate": 4.105263157894737e-07, | |
| "loss": 0.4511, | |
| "step": 1961 | |
| }, | |
| { | |
| "epoch": 3.0324574961360122, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.4024, | |
| "step": 1962 | |
| }, | |
| { | |
| "epoch": 3.034003091190108, | |
| "grad_norm": 12.875, | |
| "learning_rate": 3.8947368421052636e-07, | |
| "loss": 0.4278, | |
| "step": 1963 | |
| }, | |
| { | |
| "epoch": 3.035548686244204, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 3.789473684210527e-07, | |
| "loss": 0.3991, | |
| "step": 1964 | |
| }, | |
| { | |
| "epoch": 3.0370942812983, | |
| "grad_norm": 9.875, | |
| "learning_rate": 3.6842105263157896e-07, | |
| "loss": 0.4595, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 3.038639876352396, | |
| "grad_norm": 8.875, | |
| "learning_rate": 3.578947368421053e-07, | |
| "loss": 0.3653, | |
| "step": 1966 | |
| }, | |
| { | |
| "epoch": 3.0401854714064913, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 3.4736842105263157e-07, | |
| "loss": 0.408, | |
| "step": 1967 | |
| }, | |
| { | |
| "epoch": 3.041731066460587, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 3.368421052631579e-07, | |
| "loss": 0.3697, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 3.043276661514683, | |
| "grad_norm": 9.0, | |
| "learning_rate": 3.263157894736842e-07, | |
| "loss": 0.4205, | |
| "step": 1969 | |
| }, | |
| { | |
| "epoch": 3.044822256568779, | |
| "grad_norm": 9.625, | |
| "learning_rate": 3.1578947368421055e-07, | |
| "loss": 0.3571, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.046367851622875, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 3.052631578947369e-07, | |
| "loss": 0.4104, | |
| "step": 1971 | |
| }, | |
| { | |
| "epoch": 3.047913446676971, | |
| "grad_norm": 9.875, | |
| "learning_rate": 2.9473684210526315e-07, | |
| "loss": 0.3788, | |
| "step": 1972 | |
| }, | |
| { | |
| "epoch": 3.0494590417310663, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.842105263157895e-07, | |
| "loss": 0.4585, | |
| "step": 1973 | |
| }, | |
| { | |
| "epoch": 3.051004636785162, | |
| "grad_norm": 10.0, | |
| "learning_rate": 2.736842105263158e-07, | |
| "loss": 0.4223, | |
| "step": 1974 | |
| }, | |
| { | |
| "epoch": 3.052550231839258, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.6315789473684213e-07, | |
| "loss": 0.3643, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 3.054095826893354, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.5263157894736846e-07, | |
| "loss": 0.4294, | |
| "step": 1976 | |
| }, | |
| { | |
| "epoch": 3.05564142194745, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 2.4210526315789473e-07, | |
| "loss": 0.3805, | |
| "step": 1977 | |
| }, | |
| { | |
| "epoch": 3.0571870170015454, | |
| "grad_norm": 9.125, | |
| "learning_rate": 2.315789473684211e-07, | |
| "loss": 0.4199, | |
| "step": 1978 | |
| }, | |
| { | |
| "epoch": 3.0587326120556413, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 2.2105263157894736e-07, | |
| "loss": 0.471, | |
| "step": 1979 | |
| }, | |
| { | |
| "epoch": 3.060278207109737, | |
| "grad_norm": 13.5, | |
| "learning_rate": 2.105263157894737e-07, | |
| "loss": 0.4051, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.061823802163833, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 0.3503, | |
| "step": 1981 | |
| }, | |
| { | |
| "epoch": 3.063369397217929, | |
| "grad_norm": 11.25, | |
| "learning_rate": 1.8947368421052634e-07, | |
| "loss": 0.3959, | |
| "step": 1982 | |
| }, | |
| { | |
| "epoch": 3.064914992272025, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.7894736842105265e-07, | |
| "loss": 0.3787, | |
| "step": 1983 | |
| }, | |
| { | |
| "epoch": 3.0664605873261204, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.6842105263157895e-07, | |
| "loss": 0.3268, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 3.0680061823802163, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.5789473684210527e-07, | |
| "loss": 0.3751, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 3.069551777434312, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.4736842105263158e-07, | |
| "loss": 0.3919, | |
| "step": 1986 | |
| }, | |
| { | |
| "epoch": 3.071097372488408, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.368421052631579e-07, | |
| "loss": 0.3795, | |
| "step": 1987 | |
| }, | |
| { | |
| "epoch": 3.072642967542504, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.2631578947368423e-07, | |
| "loss": 0.3608, | |
| "step": 1988 | |
| }, | |
| { | |
| "epoch": 3.0741885625966, | |
| "grad_norm": 13.5, | |
| "learning_rate": 1.1578947368421054e-07, | |
| "loss": 0.4163, | |
| "step": 1989 | |
| }, | |
| { | |
| "epoch": 3.0757341576506954, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.0526315789473685e-07, | |
| "loss": 0.4502, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0772797527047913, | |
| "grad_norm": 10.125, | |
| "learning_rate": 9.473684210526317e-08, | |
| "loss": 0.4118, | |
| "step": 1991 | |
| }, | |
| { | |
| "epoch": 3.078825347758887, | |
| "grad_norm": 11.25, | |
| "learning_rate": 8.421052631578947e-08, | |
| "loss": 0.3723, | |
| "step": 1992 | |
| }, | |
| { | |
| "epoch": 3.080370942812983, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 7.368421052631579e-08, | |
| "loss": 0.4009, | |
| "step": 1993 | |
| }, | |
| { | |
| "epoch": 3.081916537867079, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 6.315789473684211e-08, | |
| "loss": 0.3519, | |
| "step": 1994 | |
| }, | |
| { | |
| "epoch": 3.0834621329211744, | |
| "grad_norm": 9.625, | |
| "learning_rate": 5.263157894736842e-08, | |
| "loss": 0.3874, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 3.0850077279752703, | |
| "grad_norm": 11.375, | |
| "learning_rate": 4.2105263157894737e-08, | |
| "loss": 0.4415, | |
| "step": 1996 | |
| }, | |
| { | |
| "epoch": 3.0865533230293662, | |
| "grad_norm": 10.0, | |
| "learning_rate": 3.157894736842106e-08, | |
| "loss": 0.3404, | |
| "step": 1997 | |
| }, | |
| { | |
| "epoch": 3.088098918083462, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 2.1052631578947368e-08, | |
| "loss": 0.3715, | |
| "step": 1998 | |
| }, | |
| { | |
| "epoch": 3.089644513137558, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.0526315789473684e-08, | |
| "loss": 0.3938, | |
| "step": 1999 | |
| }, | |
| { | |
| "epoch": 3.091190108191654, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 0.0, | |
| "loss": 0.3763, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |