url stringlengths 56 57 | repository_url stringclasses 1
value | labels_url stringlengths 70 71 | comments_url stringlengths 65 66 | events_url stringlengths 63 64 | html_url stringlengths 46 47 | id int64 1.88B 2.91B | node_id stringlengths 18 18 | number int64 906 2.42k | title stringlengths 3 380 | user dict | labels listlengths 0 3 | state stringclasses 2
values | locked bool 1
class | assignee dict | assignees listlengths 0 1 | milestone null | comments int64 0 42 | created_at timestamp[s] | updated_at timestamp[s] | closed_at timestamp[s] | author_association stringclasses 4
values | sub_issues_summary dict | active_lock_reason null | draft bool 0
classes | pull_request dict | body stringlengths 4 45.9k ⌀ | closed_by dict | reactions dict | timeline_url stringlengths 65 66 | performed_via_github_app null | state_reason stringclasses 3
values | is_pull_request bool 1
class |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/huggingface/peft/issues/1731 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1731/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1731/comments | https://api.github.com/repos/huggingface/peft/issues/1731/events | https://github.com/huggingface/peft/issues/1731 | 2,295,776,967 | I_kwDOIf9iDM6I1sbH | 1,731 | DoRA training in distributed setting | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | [] | closed | false | null | [] | null | 1 | 2024-05-14T15:23:06 | 2024-06-23T15:03:25 | 2024-06-23T15:03:25 | MEMBER | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
-
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
-
### Expected behavior
As reported by @w... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1731/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1731/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1730 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1730/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1730/comments | https://api.github.com/repos/huggingface/peft/issues/1730/events | https://github.com/huggingface/peft/issues/1730 | 2,295,510,601 | I_kwDOIf9iDM6I0rZJ | 1,730 | Delete certain layers from PEFT model. | {
"login": "whr819987540",
"id": 68956118,
"node_id": "MDQ6VXNlcjY4OTU2MTE4",
"avatar_url": "https://avatars.githubusercontent.com/u/68956118?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/whr819987540",
"html_url": "https://github.com/whr819987540",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 6 | 2024-05-14T13:46:54 | 2024-06-05T13:22:20 | 2024-06-05T13:22:20 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
python==3.7.12
PEFT==0.3.0
### Who can help?
@BenjaminBossan
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [X] My own task or dataset (give details below)
### Reproduction
For compatibil... | {
"login": "whr819987540",
"id": 68956118,
"node_id": "MDQ6VXNlcjY4OTU2MTE4",
"avatar_url": "https://avatars.githubusercontent.com/u/68956118?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/whr819987540",
"html_url": "https://github.com/whr819987540",
"followers_url": "https://api.github.c... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1730/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1730/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1729 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1729/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1729/comments | https://api.github.com/repos/huggingface/peft/issues/1729/events | https://github.com/huggingface/peft/issues/1729 | 2,294,331,377 | I_kwDOIf9iDM6IwLfx | 1,729 | Trainer.train() giving me Key Error: [random number] | {
"login": "fishroll23",
"id": 152573676,
"node_id": "U_kgDOCRgW7A",
"avatar_url": "https://avatars.githubusercontent.com/u/152573676?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fishroll23",
"html_url": "https://github.com/fishroll23",
"followers_url": "https://api.github.com/users/fis... | [] | closed | false | null | [] | null | 3 | 2024-05-14T03:53:10 | 2024-05-14T15:11:31 | 2024-05-14T15:11:31 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft == 0.10.0
transformers==4.40.2
python 3.10.11
### Code
```
from peft import LoraConfig, TaskType, get_peft_model
from transformers import AutoTokenizer, AutoModelForSequenceClassification, TrainingArguments, Trainer
from sklearn.metrics import accuracy_score, precision_recall_fscore_sup... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1729/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1729/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1728 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1728/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1728/comments | https://api.github.com/repos/huggingface/peft/issues/1728/events | https://github.com/huggingface/peft/issues/1728 | 2,293,086,874 | I_kwDOIf9iDM6Irbqa | 1,728 | Initialization for LoRA weights A and B initialized | {
"login": "sanaullah-06",
"id": 76161635,
"node_id": "MDQ6VXNlcjc2MTYxNjM1",
"avatar_url": "https://avatars.githubusercontent.com/u/76161635?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sanaullah-06",
"html_url": "https://github.com/sanaullah-06",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 2 | 2024-05-13T15:01:10 | 2024-06-13T09:58:27 | 2024-06-13T09:58:27 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
The comment and code are contradictory please anyone explain it to me.
### Who can help?
@BenjaminBossan
### Information
- [ ] The official example scripts
- [ ] My own modi... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1728/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1728/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1726 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1726/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1726/comments | https://api.github.com/repos/huggingface/peft/issues/1726/events | https://github.com/huggingface/peft/issues/1726 | 2,290,799,884 | I_kwDOIf9iDM6IitUM | 1,726 | Add support for OpenELM LoRA fine-tuning | {
"login": "RonanKMcGovern",
"id": 78278410,
"node_id": "MDQ6VXNlcjc4Mjc4NDEw",
"avatar_url": "https://avatars.githubusercontent.com/u/78278410?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RonanKMcGovern",
"html_url": "https://github.com/RonanKMcGovern",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | null | 2 | 2024-05-11T09:12:29 | 2024-05-13T21:53:36 | 2024-05-13T21:53:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Allow for fine-tuning with LoRA peft.
### Motivation
Full-fine-tuning is difficult for these models and does not generate great results. Perhaps LoRA would provide useful regularization/smoothing and improved results.
### Your contribution
Right now, when I run LoRA fine-tuning, it appears tha... | {
"login": "RonanKMcGovern",
"id": 78278410,
"node_id": "MDQ6VXNlcjc4Mjc4NDEw",
"avatar_url": "https://avatars.githubusercontent.com/u/78278410?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RonanKMcGovern",
"html_url": "https://github.com/RonanKMcGovern",
"followers_url": "https://api.gi... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1726/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1726/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1721 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1721/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1721/comments | https://api.github.com/repos/huggingface/peft/issues/1721/events | https://github.com/huggingface/peft/issues/1721 | 2,288,361,246 | I_kwDOIf9iDM6IZZ8e | 1,721 | TypeError: unsupported operand type(s) for *: 'Parameter' and 'NoneType' | {
"login": "misonsky",
"id": 77843536,
"node_id": "MDQ6VXNlcjc3ODQzNTM2",
"avatar_url": "https://avatars.githubusercontent.com/u/77843536?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/misonsky",
"html_url": "https://github.com/misonsky",
"followers_url": "https://api.github.com/users/mis... | [] | closed | false | null | [] | null | 2 | 2024-05-09T19:59:40 | 2024-06-17T15:03:32 | 2024-06-17T15:03:32 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
### Adalora
```
def update_ipt(self, model):
# Update the sensitivity and uncertainty for every weight
for n, p in model.named_parameters():
if "lora_" in n and self.adapter_name in n:
if n not in self.ipt:
self.ipt[n] = torch.zeros_like(p)
... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1721/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1721/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1720 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1720/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1720/comments | https://api.github.com/repos/huggingface/peft/issues/1720/events | https://github.com/huggingface/peft/issues/1720 | 2,287,857,646 | I_kwDOIf9iDM6IXe_u | 1,720 | RuntimeError: only Tensors of floating point dtype can require gradients for QLoRA since transformers 4.40 | {
"login": "dipanjanS",
"id": 3448263,
"node_id": "MDQ6VXNlcjM0NDgyNjM=",
"avatar_url": "https://avatars.githubusercontent.com/u/3448263?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dipanjanS",
"html_url": "https://github.com/dipanjanS",
"followers_url": "https://api.github.com/users/di... | [] | closed | false | null | [] | null | 7 | 2024-05-09T14:51:11 | 2024-07-12T14:43:46 | 2024-06-17T15:03:34 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```
transformers-4.40.1
peft-0.10.0
accelerate-0.30.0
bitsandbytes-0.43.1
evaluate-0.4.2
```
### Who can help?
@pacman100
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` fold... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1720/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1720/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1718 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1718/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1718/comments | https://api.github.com/repos/huggingface/peft/issues/1718/events | https://github.com/huggingface/peft/issues/1718 | 2,286,921,161 | I_kwDOIf9iDM6IT6XJ | 1,718 | Unable to Install peft==0.7.0.dev0 | {
"login": "AnonymXXXXX",
"id": 142968218,
"node_id": "U_kgDOCIWFmg",
"avatar_url": "https://avatars.githubusercontent.com/u/142968218?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AnonymXXXXX",
"html_url": "https://github.com/AnonymXXXXX",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 2 | 2024-05-09T05:27:58 | 2024-05-11T13:30:44 | 2024-05-11T13:30:44 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```bash
ERROR: Could not find a version that satisfies the requirement peft==0.7.0.dev0 (from versions: 0.0.1, 0.0.2, 0.1.0, 0.2.0, 0.3.0, 0.4.0, 0.5.0, 0.6.0, 0.6.1, 0.6.2, 0.7.0, 0.7.1, 0.8.0, 0.8.1, 0.8.2, 0.9.0, 0.10.0)
ERROR: No matching distribution found for peft==0.7.0.dev0
```
### Who ca... | {
"login": "AnonymXXXXX",
"id": 142968218,
"node_id": "U_kgDOCIWFmg",
"avatar_url": "https://avatars.githubusercontent.com/u/142968218?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AnonymXXXXX",
"html_url": "https://github.com/AnonymXXXXX",
"followers_url": "https://api.github.com/users/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1718/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1718/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1716 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1716/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1716/comments | https://api.github.com/repos/huggingface/peft/issues/1716/events | https://github.com/huggingface/peft/issues/1716 | 2,284,844,886 | I_kwDOIf9iDM6IL_dW | 1,716 | GPU Allocation Issue (QLoRa + Llama3-8B-IT) | {
"login": "DONGRYEOLLEE1",
"id": 82855597,
"node_id": "MDQ6VXNlcjgyODU1NTk3",
"avatar_url": "https://avatars.githubusercontent.com/u/82855597?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/DONGRYEOLLEE1",
"html_url": "https://github.com/DONGRYEOLLEE1",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | null | 1 | 2024-05-08T07:04:46 | 2024-05-28T04:28:28 | 2024-05-28T04:28:28 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft: 0.10.1.dev0
accelerate: 0.30.0
bitsandbytes: 0.43.1
transformers: 4.39.3
GPU: A6000 * 2 ( 96GB )
nvidia-driver version: 535.171.04
cuda: 11.8
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ]... | {
"login": "DONGRYEOLLEE1",
"id": 82855597,
"node_id": "MDQ6VXNlcjgyODU1NTk3",
"avatar_url": "https://avatars.githubusercontent.com/u/82855597?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/DONGRYEOLLEE1",
"html_url": "https://github.com/DONGRYEOLLEE1",
"followers_url": "https://api.githu... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1716/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1716/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1715 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1715/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1715/comments | https://api.github.com/repos/huggingface/peft/issues/1715/events | https://github.com/huggingface/peft/issues/1715 | 2,283,880,202 | I_kwDOIf9iDM6IIT8K | 1,715 | eval_loss showing Nan but train_loss decreases and goes to NaN after couple of steps while fine tuning gemma model with additional vocab | {
"login": "sidtandon2014",
"id": 6470596,
"node_id": "MDQ6VXNlcjY0NzA1OTY=",
"avatar_url": "https://avatars.githubusercontent.com/u/6470596?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sidtandon2014",
"html_url": "https://github.com/sidtandon2014",
"followers_url": "https://api.github.... | [] | closed | false | null | [] | null | 2 | 2024-05-07T17:28:49 | 2024-06-15T15:04:27 | 2024-06-15T15:04:26 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
I am trying to fine tune gemma 7b model in 4 bit with additional vocab and using following configuration, but getting NaN in train and eval loss. Though train loss first decreases for couple of steps and then turn to NaN
```
bnb_config = BitsAndBytesConfig(
load_in_4bit=True,
bnb_4bit_q... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1715/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1715/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1712 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1712/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1712/comments | https://api.github.com/repos/huggingface/peft/issues/1712/events | https://github.com/huggingface/peft/issues/1712 | 2,280,135,822 | I_kwDOIf9iDM6H6ByO | 1,712 | how to finetune whisper model with 'initial_prompt' | {
"login": "ben-8878",
"id": 38179632,
"node_id": "MDQ6VXNlcjM4MTc5NjMy",
"avatar_url": "https://avatars.githubusercontent.com/u/38179632?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ben-8878",
"html_url": "https://github.com/ben-8878",
"followers_url": "https://api.github.com/users/ben... | [] | closed | false | null | [] | null | 4 | 2024-05-06T06:28:20 | 2024-06-13T15:03:43 | 2024-06-13T15:03:43 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | when use 'initial_prompt', the decoding result of finetuning with my data on whisper model v2 is bad, on the contrary, the result is good.
however, when use 'initial_prompt' the decoding result of based whisper model v2 is also good, so it means If want to use 'initial_prompt' during decoding , must add it when t... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1712/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1712/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1711 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1711/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1711/comments | https://api.github.com/repos/huggingface/peft/issues/1711/events | https://github.com/huggingface/peft/issues/1711 | 2,279,525,553 | I_kwDOIf9iDM6H3syx | 1,711 | Can peft support ColumnParallelLinear? | {
"login": "wjn1996",
"id": 24487364,
"node_id": "MDQ6VXNlcjI0NDg3MzY0",
"avatar_url": "https://avatars.githubusercontent.com/u/24487364?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/wjn1996",
"html_url": "https://github.com/wjn1996",
"followers_url": "https://api.github.com/users/wjn199... | [] | closed | false | null | [] | null | 2 | 2024-05-05T13:09:44 | 2024-06-13T15:03:45 | 2024-06-13T15:03:45 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
I have a model, and the architecture has xxxParallel attributes, which are used for parallel inference:
```
BaichuanForCausalLM(
(model): BaiChuanModel(
(embed_tokens): VocabParallelEmbedding()
(layers): ModuleList(
(0-31): 32 x BaiChuanDecoderLayer(
(self_attn): BaiChuan... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1711/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1711/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1710 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1710/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1710/comments | https://api.github.com/repos/huggingface/peft/issues/1710/events | https://github.com/huggingface/peft/issues/1710 | 2,278,962,372 | I_kwDOIf9iDM6H1jTE | 1,710 | PeftModel failing to load after finetuning. Size Mismatch Error | {
"login": "sunxiaojie99",
"id": 41667783,
"node_id": "MDQ6VXNlcjQxNjY3Nzgz",
"avatar_url": "https://avatars.githubusercontent.com/u/41667783?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sunxiaojie99",
"html_url": "https://github.com/sunxiaojie99",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 5 | 2024-05-04T12:03:46 | 2024-09-20T10:32:30 | 2024-06-11T15:03:32 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
I trained the `Mistral-7B-Instruct-v0-1` model using LoRa and deepspeed
- Peft version : 0.10.0
- transformers version : 4.40.0
Code to load model after fine-tuning:
```
base_model = cls.TRANSFORMER_CLS.from_pretrained(model_name_or_path, **hf_kwargs)
lora_config = LoraConfig.from_pretrained(lo... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1710/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1710/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1709 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1709/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1709/comments | https://api.github.com/repos/huggingface/peft/issues/1709/events | https://github.com/huggingface/peft/issues/1709 | 2,278,794,632 | I_kwDOIf9iDM6H06WI | 1,709 | Ignore keys for modules to save | {
"login": "zankner",
"id": 39166683,
"node_id": "MDQ6VXNlcjM5MTY2Njgz",
"avatar_url": "https://avatars.githubusercontent.com/u/39166683?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zankner",
"html_url": "https://github.com/zankner",
"followers_url": "https://api.github.com/users/zankne... | [] | closed | false | null | [] | null | 2 | 2024-05-04T06:44:22 | 2024-06-12T15:03:36 | 2024-06-12T15:03:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
It would be nice if you could specify a regex pattern for layer names to ignore from target modules.
### Motivation
I have a base model that I apply lora to, but then I also add a few extra attention layers at the end of the model. I don't want lora to target these new decoder layers.
### Your c... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1709/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1709/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1708 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1708/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1708/comments | https://api.github.com/repos/huggingface/peft/issues/1708/events | https://github.com/huggingface/peft/issues/1708 | 2,278,171,570 | I_kwDOIf9iDM6HyiOy | 1,708 | OOM with Phi-3-mini (3.8B) on 83.5GB RAM due to LoftQ | {
"login": "adamamer20",
"id": 136176500,
"node_id": "U_kgDOCB3jdA",
"avatar_url": "https://avatars.githubusercontent.com/u/136176500?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/adamamer20",
"html_url": "https://github.com/adamamer20",
"followers_url": "https://api.github.com/users/ada... | [] | closed | false | null | [] | null | 4 | 2024-05-03T18:01:12 | 2024-05-10T09:18:21 | 2024-05-10T09:18:21 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
System: Linux-6.1.58+-x86_64-with-glibc2.35 / Google Colab
peft: 0.10.0
transformers: 4.40.1
accelerate: 0.30.0
Python 3.10.12
RAM: 83.5 GB
GPU: A100 40GB
CPU: Intel(R) Xeon(R) CPU @ 2.20GHz
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [... | {
"login": "adamamer20",
"id": 136176500,
"node_id": "U_kgDOCB3jdA",
"avatar_url": "https://avatars.githubusercontent.com/u/136176500?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/adamamer20",
"html_url": "https://github.com/adamamer20",
"followers_url": "https://api.github.com/users/ada... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1708/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1708/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1707 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1707/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1707/comments | https://api.github.com/repos/huggingface/peft/issues/1707/events | https://github.com/huggingface/peft/issues/1707 | 2,278,099,709 | I_kwDOIf9iDM6HyQr9 | 1,707 | model merge_and_unload do not support layer_replication | {
"login": "CrazyBoyM",
"id": 35400185,
"node_id": "MDQ6VXNlcjM1NDAwMTg1",
"avatar_url": "https://avatars.githubusercontent.com/u/35400185?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/CrazyBoyM",
"html_url": "https://github.com/CrazyBoyM",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 6 | 2024-05-03T17:12:04 | 2024-06-11T15:03:34 | 2024-06-11T15:03:34 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
when you trained a model with layer_replication in LoRAConfig，you will find that can not be merge to the base model in a right way
### Who can help?
_No response_
### Information
- [x] The official example scripts
- [x] My own modified scripts
### Tasks
- [x] An officially supported task in the `... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1707/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1707/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1704 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1704/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1704/comments | https://api.github.com/repos/huggingface/peft/issues/1704/events | https://github.com/huggingface/peft/issues/1704 | 2,275,428,453 | I_kwDOIf9iDM6HoEhl | 1,704 | Support merge_and_unload for IA3 Adapters with 4-bit and 8bit Quantization models | {
"login": "Abdullah-kwl",
"id": 101694671,
"node_id": "U_kgDOBg-8zw",
"avatar_url": "https://avatars.githubusercontent.com/u/101694671?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Abdullah-kwl",
"html_url": "https://github.com/Abdullah-kwl",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 1 | 2024-05-02T12:33:02 | 2024-06-09T15:03:35 | 2024-06-09T15:03:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Enable merge_and_unload functionality with ia3 adapters loaded with 4-bit and 8-bit quantization model. Currently, merging fails with an error "Cannot merge ia3 layers when the model is loaded in 4-bit mode"
![image](https://github.com/huggingface/peft/assets/101694671/a7071a3f-9b7f-49c5-aafc-0f7... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1704/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1704/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1703 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1703/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1703/comments | https://api.github.com/repos/huggingface/peft/issues/1703/events | https://github.com/huggingface/peft/issues/1703 | 2,274,793,541 | I_kwDOIf9iDM6HlphF | 1,703 | OOM error while QLoRA+Deepspeed fine tuning of Llama3-70B model on 4xA100-40GB gpus | {
"login": "hrushikesh198",
"id": 6188036,
"node_id": "MDQ6VXNlcjYxODgwMzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/6188036?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hrushikesh198",
"html_url": "https://github.com/hrushikesh198",
"followers_url": "https://api.github.... | [] | closed | false | null | [] | null | 3 | 2024-05-02T07:16:08 | 2024-06-09T15:03:36 | 2024-06-09T15:03:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```
OS Ubuntu
GPUS 4XA100-40GB
Python 3.10.14
accelerate 0.29.3
bitsandbytes 0.43.1
deepspeed 0.14.2
peft 0.10.0
transformers 4.40.1
trl ... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1703/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 1
} | https://api.github.com/repos/huggingface/peft/issues/1703/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1702 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1702/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1702/comments | https://api.github.com/repos/huggingface/peft/issues/1702/events | https://github.com/huggingface/peft/issues/1702 | 2,273,933,704 | I_kwDOIf9iDM6HiXmI | 1,702 | The FSDP example fails to run with "ValueError: Must flatten tensors with uniform requires_grad when use_orig_params=False" | {
"login": "xiangxu-google",
"id": 117880274,
"node_id": "U_kgDOBwa10g",
"avatar_url": "https://avatars.githubusercontent.com/u/117880274?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/xiangxu-google",
"html_url": "https://github.com/xiangxu-google",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 2 | 2024-05-01T18:02:00 | 2024-05-02T00:32:48 | 2024-05-02T00:32:48 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
pytorch version: 2.2.2
peft version: 0.10.0
GPUs: 8 A100-40G
### Who can help?
_No response_
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give ... | {
"login": "xiangxu-google",
"id": 117880274,
"node_id": "U_kgDOBwa10g",
"avatar_url": "https://avatars.githubusercontent.com/u/117880274?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/xiangxu-google",
"html_url": "https://github.com/xiangxu-google",
"followers_url": "https://api.github.c... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1702/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1702/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1700 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1700/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1700/comments | https://api.github.com/repos/huggingface/peft/issues/1700/events | https://github.com/huggingface/peft/issues/1700 | 2,272,407,652 | I_kwDOIf9iDM6HcjBk | 1,700 | Error while loading PEFT lora model | {
"login": "Zuhashaik",
"id": 104340147,
"node_id": "U_kgDOBjgasw",
"avatar_url": "https://avatars.githubusercontent.com/u/104340147?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Zuhashaik",
"html_url": "https://github.com/Zuhashaik",
"followers_url": "https://api.github.com/users/Zuhash... | [] | closed | false | null | [] | null | 4 | 2024-04-30T20:36:17 | 2024-05-02T09:29:53 | 2024-05-02T09:27:53 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Trained model using this lora config:
```
model.resize_token_embeddings(len(tokenizer))
Embedding(33004, 4096)
lora_alpha = 2048
lora_dropout = 0.3
lora_r = 1024
target_modules=[
# "up_proj",
"o_proj",
"v_proj",
"gate_proj",
"q_proj",
# "down_proj",
"k_proj"
]
modules_to_... | {
"login": "Zuhashaik",
"id": 104340147,
"node_id": "U_kgDOBjgasw",
"avatar_url": "https://avatars.githubusercontent.com/u/104340147?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Zuhashaik",
"html_url": "https://github.com/Zuhashaik",
"followers_url": "https://api.github.com/users/Zuhash... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1700/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1700/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1699 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1699/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1699/comments | https://api.github.com/repos/huggingface/peft/issues/1699/events | https://github.com/huggingface/peft/issues/1699 | 2,271,299,972 | I_kwDOIf9iDM6HYUmE | 1,699 | Outdated utility function: No attribute get_module_class_from_name in FullyShardedDataParallelPlugin | {
"login": "Xirid",
"id": 36632599,
"node_id": "MDQ6VXNlcjM2NjMyNTk5",
"avatar_url": "https://avatars.githubusercontent.com/u/36632599?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Xirid",
"html_url": "https://github.com/Xirid",
"followers_url": "https://api.github.com/users/Xirid/follow... | [] | closed | false | null | [] | null | 1 | 2024-04-30T12:10:55 | 2024-04-30T12:42:47 | 2024-04-30T12:42:47 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
Accelerate & peft from main
### Who can help?
_No response_
### Information
- [x] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
Should happen wit... | {
"login": "Xirid",
"id": 36632599,
"node_id": "MDQ6VXNlcjM2NjMyNTk5",
"avatar_url": "https://avatars.githubusercontent.com/u/36632599?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Xirid",
"html_url": "https://github.com/Xirid",
"followers_url": "https://api.github.com/users/Xirid/follow... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1699/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1699/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1693 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1693/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1693/comments | https://api.github.com/repos/huggingface/peft/issues/1693/events | https://github.com/huggingface/peft/issues/1693 | 2,270,693,771 | I_kwDOIf9iDM6HWAmL | 1,693 | How to convert a loha safetensor trained from diffusers to webui format | {
"login": "JIAOJIAYUASD",
"id": 61881733,
"node_id": "MDQ6VXNlcjYxODgxNzMz",
"avatar_url": "https://avatars.githubusercontent.com/u/61881733?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/JIAOJIAYUASD",
"html_url": "https://github.com/JIAOJIAYUASD",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 2 | 2024-04-30T07:17:48 | 2024-06-08T15:03:44 | 2024-06-08T15:03:44 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello, when I finetune SDXL (actually that is InstantID) with PEFT method, I use lora、loha and lokr for PEFT in [diffuser](https://github.com/huggingface/diffusers).
I have a question, how to convert a loha safetensor trained from diffusers to webui format？
In the training process:
the loading way:
`peft_config =... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1693/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1693/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1692 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1692/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1692/comments | https://api.github.com/repos/huggingface/peft/issues/1692/events | https://github.com/huggingface/peft/issues/1692 | 2,270,516,089 | I_kwDOIf9iDM6HVVN5 | 1,692 | DoRA uses lots of GPU VRAM due to fp32 upcasting | {
"login": "rationalism",
"id": 813306,
"node_id": "MDQ6VXNlcjgxMzMwNg==",
"avatar_url": "https://avatars.githubusercontent.com/u/813306?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rationalism",
"html_url": "https://github.com/rationalism",
"followers_url": "https://api.github.com/user... | [] | closed | false | null | [] | null | 7 | 2024-04-30T05:07:43 | 2024-06-23T15:03:26 | 2024-06-23T15:03:26 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft 0.10.0, transformers 4.40.1, Python 3.10 on Ubuntu 22.04
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1692/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1692/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1689 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1689/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1689/comments | https://api.github.com/repos/huggingface/peft/issues/1689/events | https://github.com/huggingface/peft/issues/1689 | 2,269,118,139 | I_kwDOIf9iDM6HP_67 | 1,689 | Saved weights differ from the original model | {
"login": "bezir",
"id": 66210642,
"node_id": "MDQ6VXNlcjY2MjEwNjQy",
"avatar_url": "https://avatars.githubusercontent.com/u/66210642?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bezir",
"html_url": "https://github.com/bezir",
"followers_url": "https://api.github.com/users/bezir/follow... | [] | closed | false | null | [] | null | 14 | 2024-04-29T13:54:12 | 2024-05-02T09:49:45 | 2024-05-02T09:49:45 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
transformers 4.40.1
peft 0.10.0
### Who can help?
@pacman100 @younesbelkada @BenjaminBossan
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [X] My own task ... | {
"login": "bezir",
"id": 66210642,
"node_id": "MDQ6VXNlcjY2MjEwNjQy",
"avatar_url": "https://avatars.githubusercontent.com/u/66210642?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bezir",
"html_url": "https://github.com/bezir",
"followers_url": "https://api.github.com/users/bezir/follow... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1689/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1689/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1688 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1688/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1688/comments | https://api.github.com/repos/huggingface/peft/issues/1688/events | https://github.com/huggingface/peft/issues/1688 | 2,268,998,455 | I_kwDOIf9iDM6HPis3 | 1,688 | Add Support for IA3 Adapters in add_weighted_adapter Method, Currently facing issue that 'IA3Model' object has no attribute 'add_weighted_adapter' | {
"login": "Abdullah-kwl",
"id": 101694671,
"node_id": "U_kgDOBg-8zw",
"avatar_url": "https://avatars.githubusercontent.com/u/101694671?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Abdullah-kwl",
"html_url": "https://github.com/Abdullah-kwl",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 9 | 2024-04-29T13:05:26 | 2024-05-17T16:59:23 | 2024-05-17T16:59:23 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
I propose adding support for IA3 adapters in the add_weighted_adapter method used within the PEFT module of the transformers library. IA3 adapters enhance model adaptability with minimal parameter increase, offering significant benefits for efficient and effective model fine-tuning across various t... | {
"login": "pacman100",
"id": 13534540,
"node_id": "MDQ6VXNlcjEzNTM0NTQw",
"avatar_url": "https://avatars.githubusercontent.com/u/13534540?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pacman100",
"html_url": "https://github.com/pacman100",
"followers_url": "https://api.github.com/users/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1688/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1688/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1696 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1696/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1696/comments | https://api.github.com/repos/huggingface/peft/issues/1696/events | https://github.com/huggingface/peft/issues/1696 | 2,270,868,039 | I_kwDOIf9iDM6HWrJH | 1,696 | Add Support for IA3 Adapters in add_weighted_adapter Method | {
"login": "Abdullah-kwl",
"id": 101694671,
"node_id": "U_kgDOBg-8zw",
"avatar_url": "https://avatars.githubusercontent.com/u/101694671?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Abdullah-kwl",
"html_url": "https://github.com/Abdullah-kwl",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 5 | 2024-04-29T13:01:42 | 2024-04-30T09:16:54 | 2024-04-30T09:16:54 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
I propose adding support for IA3 adapters in the add_weighted_adapter method used within the PEFT module of the transformers library. IA3 adapters enhance model adaptability with minimal parameter increase, offering significant benefits for efficient and effective model fine-tuning across various t... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1696/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1696/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1685 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1685/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1685/comments | https://api.github.com/repos/huggingface/peft/issues/1685/events | https://github.com/huggingface/peft/issues/1685 | 2,267,537,805 | I_kwDOIf9iDM6HJ-GN | 1,685 | Example for Prompt-based methods fails with `expected sequence of length` mismatch | {
"login": "jpodivin",
"id": 66251151,
"node_id": "MDQ6VXNlcjY2MjUxMTUx",
"avatar_url": "https://avatars.githubusercontent.com/u/66251151?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jpodivin",
"html_url": "https://github.com/jpodivin",
"followers_url": "https://api.github.com/users/jpo... | [] | closed | false | null | [] | null | 0 | 2024-04-28T12:27:21 | 2024-04-29T14:23:47 | 2024-04-29T14:23:47 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft-0.10.0, datasets-2.19.0, transformers-4.40.1, Python 3.11.9
### Who can help?
_No response_
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details be... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1685/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1685/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1678 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1678/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1678/comments | https://api.github.com/repos/huggingface/peft/issues/1678/events | https://github.com/huggingface/peft/issues/1678 | 2,262,085,263 | I_kwDOIf9iDM6G1K6P | 1,678 | Loading trained peft model results in random adapter weights each time | {
"login": "ambisinister",
"id": 16823221,
"node_id": "MDQ6VXNlcjE2ODIzMjIx",
"avatar_url": "https://avatars.githubusercontent.com/u/16823221?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ambisinister",
"html_url": "https://github.com/ambisinister",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 4 | 2024-04-24T20:00:28 | 2024-04-26T19:14:27 | 2024-04-26T19:14:26 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
- peft==0.10.0
- transformers==4.28.2
- accelerate==0.28.0
python 3.9.5
ubuntu 20.05.6
### Who can help?
_No response_
### Information
- [X] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [X]... | {
"login": "ambisinister",
"id": 16823221,
"node_id": "MDQ6VXNlcjE2ODIzMjIx",
"avatar_url": "https://avatars.githubusercontent.com/u/16823221?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ambisinister",
"html_url": "https://github.com/ambisinister",
"followers_url": "https://api.github.c... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1678/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1678/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1677 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1677/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1677/comments | https://api.github.com/repos/huggingface/peft/issues/1677/events | https://github.com/huggingface/peft/issues/1677 | 2,262,067,229 | I_kwDOIf9iDM6G1Ggd | 1,677 | DoRA support for Embedding | {
"login": "ShayekhBinIslam",
"id": 38907457,
"node_id": "MDQ6VXNlcjM4OTA3NDU3",
"avatar_url": "https://avatars.githubusercontent.com/u/38907457?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ShayekhBinIslam",
"html_url": "https://github.com/ShayekhBinIslam",
"followers_url": "https://api... | [
{
"id": 7219265350,
"node_id": "LA_kwDOIf9iDM8AAAABrk0_Rg",
"url": "https://api.github.com/repos/huggingface/peft/labels/contributions-welcome",
"name": "contributions-welcome",
"color": "F2AD28",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | 5 | 2024-04-24T19:48:14 | 2024-08-23T18:20:43 | 2024-08-23T18:20:43 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Supporting DoRA with Embedding to allow seamless integration with the LoRA framework.
### Motivation
I need to finetune a model where additional tokens have to be trained. But getting the "Embedding does not support DoRA yet, please set it to False" error.
### Your contribution
No | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1677/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1677/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1676 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1676/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1676/comments | https://api.github.com/repos/huggingface/peft/issues/1676/events | https://github.com/huggingface/peft/issues/1676 | 2,260,983,101 | I_kwDOIf9iDM6Gw909 | 1,676 | Clarification needed on Adapter Heads in PEFT | {
"login": "lenglaender",
"id": 77012866,
"node_id": "MDQ6VXNlcjc3MDEyODY2",
"avatar_url": "https://avatars.githubusercontent.com/u/77012866?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lenglaender",
"html_url": "https://github.com/lenglaender",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | null | 2 | 2024-04-24T10:41:36 | 2024-04-30T09:34:54 | 2024-04-30T09:34:53 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I have questions regarding the behaviour of adapter heads when activating and merging LoRA adapters. I couldn't find any information in the documentation on how PEFT handles heads.
I have trained and loaded multiple LoRA adapters for classification and language modelling tasks. Each classification adapter has learne... | {
"login": "lenglaender",
"id": 77012866,
"node_id": "MDQ6VXNlcjc3MDEyODY2",
"avatar_url": "https://avatars.githubusercontent.com/u/77012866?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lenglaender",
"html_url": "https://github.com/lenglaender",
"followers_url": "https://api.github.com/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1676/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1676/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1674 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1674/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1674/comments | https://api.github.com/repos/huggingface/peft/issues/1674/events | https://github.com/huggingface/peft/issues/1674 | 2,260,752,997 | I_kwDOIf9iDM6GwFpl | 1,674 | [FSDP+QLoRA] ValueError: Expected a cuda device, but got: cpu | {
"login": "iseesaw",
"id": 31267864,
"node_id": "MDQ6VXNlcjMxMjY3ODY0",
"avatar_url": "https://avatars.githubusercontent.com/u/31267864?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/iseesaw",
"html_url": "https://github.com/iseesaw",
"followers_url": "https://api.github.com/users/iseesa... | [] | closed | false | null | [] | null | 7 | 2024-04-24T08:47:41 | 2024-05-31T14:58:51 | 2024-05-14T15:10:24 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
pip list
```
accelerate 0.29.3
bitsandbytes 0.43.1
datasets 2.14.6
huggingface-hub 0.20.3
llama-recipes 0.0.1
peft 0.10.0
safetensors 0.4.2
tokenizers 0.19.1
torch ... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1674/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1674/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1673 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1673/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1673/comments | https://api.github.com/repos/huggingface/peft/issues/1673/events | https://github.com/huggingface/peft/issues/1673 | 2,260,617,561 | I_kwDOIf9iDM6GvklZ | 1,673 | How to set Lora_dropout=0 when loading trained peft model for inference? | {
"login": "flyliu2017",
"id": 34392090,
"node_id": "MDQ6VXNlcjM0MzkyMDkw",
"avatar_url": "https://avatars.githubusercontent.com/u/34392090?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/flyliu2017",
"html_url": "https://github.com/flyliu2017",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 2 | 2024-04-24T07:47:19 | 2024-05-10T02:22:17 | 2024-05-10T02:22:17 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft==0.10.0
transformers==4.39.3
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
```pytho... | {
"login": "flyliu2017",
"id": 34392090,
"node_id": "MDQ6VXNlcjM0MzkyMDkw",
"avatar_url": "https://avatars.githubusercontent.com/u/34392090?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/flyliu2017",
"html_url": "https://github.com/flyliu2017",
"followers_url": "https://api.github.com/use... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1673/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1673/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1672 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1672/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1672/comments | https://api.github.com/repos/huggingface/peft/issues/1672/events | https://github.com/huggingface/peft/issues/1672 | 2,259,502,756 | I_kwDOIf9iDM6GrUak | 1,672 | Using PEFT causes model to not predict EOS | {
"login": "Km3888",
"id": 32725461,
"node_id": "MDQ6VXNlcjMyNzI1NDYx",
"avatar_url": "https://avatars.githubusercontent.com/u/32725461?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Km3888",
"html_url": "https://github.com/Km3888",
"followers_url": "https://api.github.com/users/Km3888/fo... | [] | closed | false | null | [] | null | 15 | 2024-04-23T18:23:50 | 2024-08-06T10:00:29 | 2024-07-01T15:03:46 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
> peft version: 0.9.0
> accelerate version: 0.27.2
> transformers version: 4.37.0
> trl version: 0.7.12.dev0
> base model: openai-community/gpt2
> hardware: 2xA100
I'm doing a LORA peft of GPT2 through [trl](https://github.com/huggingface/trl) and have noticed that my trained model assigns very... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1672/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1672/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1670 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1670/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1670/comments | https://api.github.com/repos/huggingface/peft/issues/1670/events | https://github.com/huggingface/peft/issues/1670 | 2,258,987,771 | I_kwDOIf9iDM6GpWr7 | 1,670 | Updating only one adapter using multi adapter | {
"login": "WoosungMyung",
"id": 115716986,
"node_id": "U_kgDOBuWzeg",
"avatar_url": "https://avatars.githubusercontent.com/u/115716986?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/WoosungMyung",
"html_url": "https://github.com/WoosungMyung",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 2 | 2024-04-23T14:10:06 | 2024-06-01T15:03:35 | 2024-06-01T15:03:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I am curious if I can update only one adapter in multi adapter siutations.
For example, I have two adapter A and B.
In forward pass, LLM + adapter A + adapter B to get Loss,
and update only adapter A using peft.
I found _mixed_batch_forward in your code, is it related to my question?
```
def _mixed_ba... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1670/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1670/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1669 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1669/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1669/comments | https://api.github.com/repos/huggingface/peft/issues/1669/events | https://github.com/huggingface/peft/issues/1669 | 2,258,374,628 | I_kwDOIf9iDM6GnA_k | 1,669 | Inquiry on Inserting P-Tuning Soft Prompts at Any Position in Input | {
"login": "Betty1202",
"id": 52363993,
"node_id": "MDQ6VXNlcjUyMzYzOTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/52363993?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Betty1202",
"html_url": "https://github.com/Betty1202",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 2 | 2024-04-23T09:20:04 | 2024-06-01T15:03:37 | 2024-06-01T15:03:37 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
None
### Who can help?
_No response_
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
None
### Expected behavior
How to inse... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1669/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1669/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1665 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1665/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1665/comments | https://api.github.com/repos/huggingface/peft/issues/1665/events | https://github.com/huggingface/peft/issues/1665 | 2,254,544,845 | I_kwDOIf9iDM6GYZ_N | 1,665 | `bitsandbytes` is imported eagerly in `peft/utils/loftq_utils.py` | {
"login": "function2-llx",
"id": 38486514,
"node_id": "MDQ6VXNlcjM4NDg2NTE0",
"avatar_url": "https://avatars.githubusercontent.com/u/38486514?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/function2-llx",
"html_url": "https://github.com/function2-llx",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | null | 5 | 2024-04-20T13:08:37 | 2024-04-25T18:35:17 | 2024-04-25T18:35:17 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | https://github.com/huggingface/peft/pull/1230 makes the import of `bitsandbytes` lazy, however, the fix seems to be imcomplete.
https://github.com/huggingface/peft/blob/5a4b9cade64bac8afdff5006ee9dd815c90b5469/src/peft/utils/loftq_utils.py#L34-L35
Perhaps there can be a test case ensuring that the CUDA context is not... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1665/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1665/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1661 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1661/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1661/comments | https://api.github.com/repos/huggingface/peft/issues/1661/events | https://github.com/huggingface/peft/issues/1661 | 2,249,685,847 | I_kwDOIf9iDM6GF3tX | 1,661 | TypeError: PeftConfig.__init__() got an unexpected keyword argument '_name_or_path' 报错怎么解决 | {
"login": "caijx168",
"id": 54843188,
"node_id": "MDQ6VXNlcjU0ODQzMTg4",
"avatar_url": "https://avatars.githubusercontent.com/u/54843188?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/caijx168",
"html_url": "https://github.com/caijx168",
"followers_url": "https://api.github.com/users/cai... | [] | closed | false | null | [] | null | 6 | 2024-04-18T03:59:34 | 2024-05-26T15:04:01 | 2024-05-26T15:04:01 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft 版本为0.10.0
### Who can help?
@pacman100 @younesbelkada @BenjaminBossan @sayakpaul @stevhliu
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1661/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1661/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1656 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1656/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1656/comments | https://api.github.com/repos/huggingface/peft/issues/1656/events | https://github.com/huggingface/peft/issues/1656 | 2,245,014,928 | I_kwDOIf9iDM6F0DWQ | 1,656 | RuntimeError: CUDA error: invalid device ordinal | {
"login": "yumath",
"id": 22947330,
"node_id": "MDQ6VXNlcjIyOTQ3MzMw",
"avatar_url": "https://avatars.githubusercontent.com/u/22947330?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yumath",
"html_url": "https://github.com/yumath",
"followers_url": "https://api.github.com/users/yumath/fo... | [] | closed | false | null | [] | null | 2 | 2024-04-16T03:25:08 | 2024-05-21T08:02:00 | 2024-05-21T08:02:00 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```
File "/home/xxx/anaconda3/envs/xxx/lib/python3.11/site-packages/accelerate/state.py", line 211, in __init__
torch.cuda.set_device(self.device)
File "/home/xxx/anaconda3/envs/xxx/lib/python3.11/site-packages/torch/cuda/__init__.py", line 350, in set_device
torch._C._cuda_setDevice(... | {
"login": "yumath",
"id": 22947330,
"node_id": "MDQ6VXNlcjIyOTQ3MzMw",
"avatar_url": "https://avatars.githubusercontent.com/u/22947330?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yumath",
"html_url": "https://github.com/yumath",
"followers_url": "https://api.github.com/users/yumath/fo... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1656/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1656/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1655 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1655/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1655/comments | https://api.github.com/repos/huggingface/peft/issues/1655/events | https://github.com/huggingface/peft/issues/1655 | 2,244,338,734 | I_kwDOIf9iDM6FxeQu | 1,655 | Learning the combination weights of pre-trained LoRA Modules | {
"login": "mahdibeit",
"id": 96784135,
"node_id": "U_kgDOBcTPBw",
"avatar_url": "https://avatars.githubusercontent.com/u/96784135?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mahdibeit",
"html_url": "https://github.com/mahdibeit",
"followers_url": "https://api.github.com/users/mahdibei... | [] | closed | false | null | [] | null | 5 | 2024-04-15T18:25:21 | 2024-05-24T15:03:50 | 2024-05-24T15:03:50 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
PEFT can combine pre-trained LoRA modules by averaging them or providing custom weights for weighted averaging. This [paper ](https://arxiv.org/abs/2402.15414) showed that learning these weights is better than naive averaging in few-shot adaption settings.
<img width="737" alt="WLoRA" src="https... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1655/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1655/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1654 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1654/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1654/comments | https://api.github.com/repos/huggingface/peft/issues/1654/events | https://github.com/huggingface/peft/issues/1654 | 2,243,875,518 | I_kwDOIf9iDM6FvtK- | 1,654 | Integrating ReFT | {
"login": "raven38",
"id": 4375159,
"node_id": "MDQ6VXNlcjQzNzUxNTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/4375159?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/raven38",
"html_url": "https://github.com/raven38",
"followers_url": "https://api.github.com/users/raven38/... | [] | closed | false | null | [] | null | 9 | 2024-04-15T14:46:11 | 2024-06-29T15:03:24 | 2024-06-29T15:03:24 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello all, and thank you for your great work!
[ReFT](https://arxiv.org/abs/2404.03592), a representation finetuning framework more parameter-efficient than popular PEFTs like LoRA, is announced earlier this month.
I implemented LoReFT, a instance of ReFT, up on the your `PEFT` library referencing to [the origi... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1654/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1654/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1651 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1651/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1651/comments | https://api.github.com/repos/huggingface/peft/issues/1651/events | https://github.com/huggingface/peft/issues/1651 | 2,242,235,271 | I_kwDOIf9iDM6FpcuH | 1,651 | Training with out Trainer class. (VRAM usage issue) | {
"login": "venzino-han",
"id": 46100518,
"node_id": "MDQ6VXNlcjQ2MTAwNTE4",
"avatar_url": "https://avatars.githubusercontent.com/u/46100518?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/venzino-han",
"html_url": "https://github.com/venzino-han",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | null | 4 | 2024-04-14T16:01:09 | 2024-04-18T15:22:46 | 2024-04-18T15:22:46 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
Python 3.10.12
transformers 4.40.0.dev0
peft 0.10.1.dev0
torch 2.1.2
I tried to apply lora with peft.
But it still using same amount of GPU VRAM.
How can I train model without using transformers.Trainer class?
### Who ... | {
"login": "venzino-han",
"id": 46100518,
"node_id": "MDQ6VXNlcjQ2MTAwNTE4",
"avatar_url": "https://avatars.githubusercontent.com/u/46100518?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/venzino-han",
"html_url": "https://github.com/venzino-han",
"followers_url": "https://api.github.com/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1651/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1651/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1650 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1650/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1650/comments | https://api.github.com/repos/huggingface/peft/issues/1650/events | https://github.com/huggingface/peft/issues/1650 | 2,241,815,787 | I_kwDOIf9iDM6Fn2Tr | 1,650 | Backward compatibility on saved config. | {
"login": "xkszltl",
"id": 5203025,
"node_id": "MDQ6VXNlcjUyMDMwMjU=",
"avatar_url": "https://avatars.githubusercontent.com/u/5203025?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/xkszltl",
"html_url": "https://github.com/xkszltl",
"followers_url": "https://api.github.com/users/xkszltl/... | [] | closed | false | null | [] | null | 5 | 2024-04-13T23:20:11 | 2024-06-17T15:03:38 | 2024-06-17T15:03:38 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Model trained on newer peft should be available on older version when possible.
### Motivation
We found model trained on peft 0.10 cannot be loaded with older peft due to unknown entry `layer_replication` in adapter config json.
This entry is never used by us so it's probably just the default va... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1650/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1650/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1649 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1649/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1649/comments | https://api.github.com/repos/huggingface/peft/issues/1649/events | https://github.com/huggingface/peft/issues/1649 | 2,241,787,531 | I_kwDOIf9iDM6FnvaL | 1,649 | model prediction differs before and after saving the model | {
"login": "IamExperimenting",
"id": 74839416,
"node_id": "MDQ6VXNlcjc0ODM5NDE2",
"avatar_url": "https://avatars.githubusercontent.com/u/74839416?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/IamExperimenting",
"html_url": "https://github.com/IamExperimenting",
"followers_url": "https://... | [] | closed | false | null | [] | null | 2 | 2024-04-13T21:29:24 | 2024-04-17T19:42:12 | 2024-04-17T19:42:11 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
accelerate verion: 0.29.1
Platform: Linux
python version: 3.10
transformer verion: 4.39.3
peft: 0.10.0
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples`... | {
"login": "IamExperimenting",
"id": 74839416,
"node_id": "MDQ6VXNlcjc0ODM5NDE2",
"avatar_url": "https://avatars.githubusercontent.com/u/74839416?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/IamExperimenting",
"html_url": "https://github.com/IamExperimenting",
"followers_url": "https://... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1649/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1649/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1647 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1647/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1647/comments | https://api.github.com/repos/huggingface/peft/issues/1647/events | https://github.com/huggingface/peft/issues/1647 | 2,241,666,797 | I_kwDOIf9iDM6FnR7t | 1,647 | Problem with merging AdaLoRA adapters | {
"login": "h4iku",
"id": 3812788,
"node_id": "MDQ6VXNlcjM4MTI3ODg=",
"avatar_url": "https://avatars.githubusercontent.com/u/3812788?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/h4iku",
"html_url": "https://github.com/h4iku",
"followers_url": "https://api.github.com/users/h4iku/follower... | [] | closed | false | null | [] | null | 1 | 2024-04-13T17:27:53 | 2024-04-29T11:09:35 | 2024-04-29T11:09:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```
peft @ git+https://github.com/huggingface/peft@c8974c5880b28a913e35f050e82402e34d181c63
accelerate==0.29.2
transformers==4.39.3
```
Ubuntu 22.04.4 LTS
Python 3.11.9
### Who can help?
@BenjaminBossan @pacman100
### Information
- [ ] The official example scripts
- [X] My own modified sc... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1647/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1647/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1646 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1646/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1646/comments | https://api.github.com/repos/huggingface/peft/issues/1646/events | https://github.com/huggingface/peft/issues/1646 | 2,241,396,984 | I_kwDOIf9iDM6FmQD4 | 1,646 | Adapters cannot be correctly loaded using PeftModel.from_pretrained() | {
"login": "HenryCai11",
"id": 24936331,
"node_id": "MDQ6VXNlcjI0OTM2MzMx",
"avatar_url": "https://avatars.githubusercontent.com/u/24936331?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/HenryCai11",
"html_url": "https://github.com/HenryCai11",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 3 | 2024-04-13T07:09:08 | 2024-05-22T15:03:42 | 2024-05-22T15:03:42 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft==0.10.0
transformers=4.39.3
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [X] My own task or dataset (give details below)
### Reproduction
```python
... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1646/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1646/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1645 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1645/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1645/comments | https://api.github.com/repos/huggingface/peft/issues/1645/events | https://github.com/huggingface/peft/issues/1645 | 2,241,254,997 | I_kwDOIf9iDM6FltZV | 1,645 | [docs] usability | {
"login": "stas00",
"id": 10676103,
"node_id": "MDQ6VXNlcjEwNjc2MTAz",
"avatar_url": "https://avatars.githubusercontent.com/u/10676103?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stas00",
"html_url": "https://github.com/stas00",
"followers_url": "https://api.github.com/users/stas00/fo... | [] | closed | false | null | [] | null | 3 | 2024-04-13T03:35:03 | 2024-06-15T15:04:31 | 2024-06-15T15:04:31 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I'd imagine the person who wrote [this](https://huggingface.co/docs/peft/main/en/accelerate/deepspeed#configuration) never checked to see how it gets rendered and I hope it's clear from the snapshot that this is not the most practical way of presenting information, since most of it is hidden:
![snapshot_726](https:/... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1645/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1645/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1644 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1644/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1644/comments | https://api.github.com/repos/huggingface/peft/issues/1644/events | https://github.com/huggingface/peft/issues/1644 | 2,241,252,705 | I_kwDOIf9iDM6Fls1h | 1,644 | [doc] issues with rendering docs in FF | {
"login": "stas00",
"id": 10676103,
"node_id": "MDQ6VXNlcjEwNjc2MTAz",
"avatar_url": "https://avatars.githubusercontent.com/u/10676103?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stas00",
"html_url": "https://github.com/stas00",
"followers_url": "https://api.github.com/users/stas00/fo... | [] | closed | false | null | [] | null | 6 | 2024-04-13T03:27:51 | 2024-05-22T15:03:44 | 2024-05-22T15:03:44 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | There are doc rendering issues in FF, at least for: https://huggingface.co/docs/peft/main/en/accelerate/deepspeed
Fine in chrome:
Broken in FF 124.0.1 (Ubuntu)
![snapshot_725](https://github.com/huggingface... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1644/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1644/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1643 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1643/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1643/comments | https://api.github.com/repos/huggingface/peft/issues/1643/events | https://github.com/huggingface/peft/issues/1643 | 2,240,213,058 | I_kwDOIf9iDM6FhvBC | 1,643 | Support EETQ QLoRA | {
"login": "thincal",
"id": 1206509,
"node_id": "MDQ6VXNlcjEyMDY1MDk=",
"avatar_url": "https://avatars.githubusercontent.com/u/1206509?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/thincal",
"html_url": "https://github.com/thincal",
"followers_url": "https://api.github.com/users/thincal/... | [] | closed | false | null | [] | null | 11 | 2024-04-12T13:56:02 | 2024-04-30T08:11:00 | 2024-04-26T08:20:19 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
EETQ: https://github.com/NetEase-FuXi/EETQ
EETQ quantization is with great performance during inference supported by TGI and LoRAX already, so that if it could be supported by finetuning that would be great for QLoRA training and inference.
### Motivation
Achieve good inference performance for... | {
"login": "younesbelkada",
"id": 49240599,
"node_id": "MDQ6VXNlcjQ5MjQwNTk5",
"avatar_url": "https://avatars.githubusercontent.com/u/49240599?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/younesbelkada",
"html_url": "https://github.com/younesbelkada",
"followers_url": "https://api.githu... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1643/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1643/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1639 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1639/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1639/comments | https://api.github.com/repos/huggingface/peft/issues/1639/events | https://github.com/huggingface/peft/issues/1639 | 2,236,624,286 | I_kwDOIf9iDM6FUC2e | 1,639 | GPU memory degradation despite offloading to CPU (on 50+ LoRAs for Stable Diffusion) | {
"login": "iuliaturc",
"id": 6214514,
"node_id": "MDQ6VXNlcjYyMTQ1MTQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/6214514?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/iuliaturc",
"html_url": "https://github.com/iuliaturc",
"followers_url": "https://api.github.com/users/iu... | [] | closed | false | null | [] | null | 23 | 2024-04-11T00:05:35 | 2024-06-24T15:03:51 | 2024-06-24T15:03:51 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
accelerate==0.28.0
diffusers==0.27.0
peft==0.9.0
transformers==4.38.2
Amazon EC2 instance (g5.2xlarge)
Deep Learning OSS Nvidia Driver AMI GPU PyTorch 2.0.1 (Amazon Linux 2)
### Who can help?
@pacman100 @younesbelkada
### Information
- [ ] The official example scripts
- [X] My o... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1639/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1639/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1638 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1638/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1638/comments | https://api.github.com/repos/huggingface/peft/issues/1638/events | https://github.com/huggingface/peft/issues/1638 | 2,235,686,293 | I_kwDOIf9iDM6FQd2V | 1,638 | Number of trainable parameters for Prefix Tuning method with CausalLM Task | {
"login": "krishnakanthnakkav2",
"id": 156657391,
"node_id": "U_kgDOCVZm7w",
"avatar_url": "https://avatars.githubusercontent.com/u/156657391?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/krishnakanthnakkav2",
"html_url": "https://github.com/krishnakanthnakkav2",
"followers_url": "https... | [] | closed | false | null | [] | null | 2 | 2024-04-10T14:03:24 | 2024-05-18T15:03:33 | 2024-05-18T15:03:33 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
I am running the experiments with the Prefix tuning on decoder-only models for `CAUSAL_LM` task.
In the prefix-tuning implementation at
https://github.com/huggingface/peft/blob/31c884e93469dd1391bb54eb1468311c38bbccac/src/peft/tuners/prefix_tuning/model.py#L72
the number of trainable paramet... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1638/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1638/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1636 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1636/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1636/comments | https://api.github.com/repos/huggingface/peft/issues/1636/events | https://github.com/huggingface/peft/issues/1636 | 2,233,585,772 | I_kwDOIf9iDM6FIdBs | 1,636 | Error running the deepspeed qlora example | {
"login": "AnirudhVIyer",
"id": 69951190,
"node_id": "MDQ6VXNlcjY5OTUxMTkw",
"avatar_url": "https://avatars.githubusercontent.com/u/69951190?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AnirudhVIyer",
"html_url": "https://github.com/AnirudhVIyer",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 2 | 2024-04-09T14:30:55 | 2024-04-11T14:41:41 | 2024-04-11T14:41:41 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
Hi,
I am trying to run a qlora finetuning experiment on deepspeed similar to the one given in the sft folder.
I use the same requirements.txt as given. However, I face an issue
_
```
Traceback (most recent call last):
File "/opt/ml/code/train_qlora.py", line 161, in <module>
model_args,... | {
"login": "AnirudhVIyer",
"id": 69951190,
"node_id": "MDQ6VXNlcjY5OTUxMTkw",
"avatar_url": "https://avatars.githubusercontent.com/u/69951190?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AnirudhVIyer",
"html_url": "https://github.com/AnirudhVIyer",
"followers_url": "https://api.github.c... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1636/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1636/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1633 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1633/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1633/comments | https://api.github.com/repos/huggingface/peft/issues/1633/events | https://github.com/huggingface/peft/issues/1633 | 2,232,560,066 | I_kwDOIf9iDM6FEinC | 1,633 | Prompt Tuning with Zero3 didn't work | {
"login": "tricktreat",
"id": 25740077,
"node_id": "MDQ6VXNlcjI1NzQwMDc3",
"avatar_url": "https://avatars.githubusercontent.com/u/25740077?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tricktreat",
"html_url": "https://github.com/tricktreat",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 4 | 2024-04-09T04:06:40 | 2024-04-10T09:34:01 | 2024-04-10T09:27:21 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft==0.9.0
### Who can help?
@younesbelkada @BenjaminBossan
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
When I train th... | {
"login": "tricktreat",
"id": 25740077,
"node_id": "MDQ6VXNlcjI1NzQwMDc3",
"avatar_url": "https://avatars.githubusercontent.com/u/25740077?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tricktreat",
"html_url": "https://github.com/tricktreat",
"followers_url": "https://api.github.com/use... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1633/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1633/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1632 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1632/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1632/comments | https://api.github.com/repos/huggingface/peft/issues/1632/events | https://github.com/huggingface/peft/issues/1632 | 2,232,176,169 | I_kwDOIf9iDM6FDE4p | 1,632 | Seeking Help/Question: Best practice for fine-tuning LLM with Lora but with addtional parameters | {
"login": "zhouhanxie",
"id": 47683426,
"node_id": "MDQ6VXNlcjQ3NjgzNDI2",
"avatar_url": "https://avatars.githubusercontent.com/u/47683426?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhouhanxie",
"html_url": "https://github.com/zhouhanxie",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 2 | 2024-04-08T22:11:32 | 2024-04-10T18:46:00 | 2024-04-10T18:46:00 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi PEFT Library, thanks for the cool project!
So I recently run into a use case where I want to fine-tune a LLM with Lora, but also some additional parameters. I'm currently using the Huggingface eco system for training. I wonder what would be the best practice for saving the additional parameters I want to have?
... | {
"login": "zhouhanxie",
"id": 47683426,
"node_id": "MDQ6VXNlcjQ3NjgzNDI2",
"avatar_url": "https://avatars.githubusercontent.com/u/47683426?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhouhanxie",
"html_url": "https://github.com/zhouhanxie",
"followers_url": "https://api.github.com/use... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1632/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1632/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1627 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1627/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1627/comments | https://api.github.com/repos/huggingface/peft/issues/1627/events | https://github.com/huggingface/peft/issues/1627 | 2,229,831,658 | I_kwDOIf9iDM6E6Ifq | 1,627 | Seek help for a more efficient way to use caching to train my model | {
"login": "whyiug",
"id": 6668626,
"node_id": "MDQ6VXNlcjY2Njg2MjY=",
"avatar_url": "https://avatars.githubusercontent.com/u/6668626?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/whyiug",
"html_url": "https://github.com/whyiug",
"followers_url": "https://api.github.com/users/whyiug/foll... | [] | closed | false | null | [] | null | 4 | 2024-04-07T16:01:22 | 2024-05-16T15:03:40 | 2024-05-16T15:03:40 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | When I was fine-tuning llama2 model using **lora**, I came across a problem.
The instruction dataset goes something like this:
`
"Here's the background to the problem... (1000 identical words)... Now answer the questions in context... (Different question, about 100 words)..." .
`
Each piece of data has the same... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1627/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1627/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1619 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1619/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1619/comments | https://api.github.com/repos/huggingface/peft/issues/1619/events | https://github.com/huggingface/peft/issues/1619 | 2,225,722,696 | I_kwDOIf9iDM6EqdVI | 1,619 | Documentation Clarification for loading PEFT models with AutoModelForCausalLM.from_pretrained | {
"login": "RonanKMcGovern",
"id": 78278410,
"node_id": "MDQ6VXNlcjc4Mjc4NDEw",
"avatar_url": "https://avatars.githubusercontent.com/u/78278410?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RonanKMcGovern",
"html_url": "https://github.com/RonanKMcGovern",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | null | 4 | 2024-04-04T14:46:55 | 2024-06-06T15:03:53 | 2024-06-06T15:03:53 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
- `transformers` version: 4.40.0.dev0
- Platform: Linux-5.4.0-150-generic-x86_64-with-glibc2.35
- Python version: 3.10.12
- Huggingface_hub version: 0.20.3
- Safetensors version: 0.4.2
- Accelerate version: 0.29.0.dev0
- Accelerate config: not found
- PyTorch version (GPU?): 2.2.2+cu121 (True... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1619/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1619/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1617 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1617/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1617/comments | https://api.github.com/repos/huggingface/peft/issues/1617/events | https://github.com/huggingface/peft/issues/1617 | 2,222,614,613 | I_kwDOIf9iDM6EemhV | 1,617 | replace_lora_weights_loftq seems not present in transformers | {
"login": "RonanKMcGovern",
"id": 78278410,
"node_id": "MDQ6VXNlcjc4Mjc4NDEw",
"avatar_url": "https://avatars.githubusercontent.com/u/78278410?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RonanKMcGovern",
"html_url": "https://github.com/RonanKMcGovern",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | null | 2 | 2024-04-03T11:02:32 | 2024-04-03T11:46:53 | 2024-04-03T11:46:53 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
transformers==4.38.1
bitsandbytes==0.42.0
peft==0.8.2
### Who can help?
@BenjaminBossan
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### ... | {
"login": "RonanKMcGovern",
"id": 78278410,
"node_id": "MDQ6VXNlcjc4Mjc4NDEw",
"avatar_url": "https://avatars.githubusercontent.com/u/78278410?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RonanKMcGovern",
"html_url": "https://github.com/RonanKMcGovern",
"followers_url": "https://api.gi... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1617/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1617/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1616 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1616/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1616/comments | https://api.github.com/repos/huggingface/peft/issues/1616/events | https://github.com/huggingface/peft/issues/1616 | 2,222,007,502 | I_kwDOIf9iDM6EcSTO | 1,616 | Support HQQ method. | {
"login": "Minami-su",
"id": 76865636,
"node_id": "MDQ6VXNlcjc2ODY1NjM2",
"avatar_url": "https://avatars.githubusercontent.com/u/76865636?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Minami-su",
"html_url": "https://github.com/Minami-su",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 11 | 2024-04-03T06:20:00 | 2024-05-03T15:11:48 | 2024-05-03T15:11:48 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | peft version 0.10.0
```
base_model=Qwen1.5-14B-Chat_llamafy-1bit
model = AutoHQQHFModel.from_quantized(base_model).half().cuda()
modules = find_all_linear_names(model)
# Add Peft
######################################################################################
config = LoraConfig(
r=lora_r,
lora_alpha=lor... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1616/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1616/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1613 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1613/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1613/comments | https://api.github.com/repos/huggingface/peft/issues/1613/events | https://github.com/huggingface/peft/issues/1613 | 2,219,888,774 | I_kwDOIf9iDM6EUNCG | 1,613 | peft v0.10 takes up too much GPU memory than v0.3.0 | {
"login": "azuryl",
"id": 23340671,
"node_id": "MDQ6VXNlcjIzMzQwNjcx",
"avatar_url": "https://avatars.githubusercontent.com/u/23340671?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/azuryl",
"html_url": "https://github.com/azuryl",
"followers_url": "https://api.github.com/users/azuryl/fo... | [] | closed | false | null | [] | null | 14 | 2024-04-02T08:28:44 | 2024-04-30T09:18:55 | 2024-04-30T09:18:55 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft v0.10
### Who can help?
@pacman100 @younesbelkada @BenjaminBossan @sayakpaul
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [X] My own task or dataset (give details bel... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1613/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1613/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1612 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1612/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1612/comments | https://api.github.com/repos/huggingface/peft/issues/1612/events | https://github.com/huggingface/peft/issues/1612 | 2,218,459,399 | I_kwDOIf9iDM6EOwEH | 1,612 | LoRA layer param.grad=None after loss.backward() using lora for mamba 2.8B | {
"login": "jinlHe",
"id": 72618505,
"node_id": "MDQ6VXNlcjcyNjE4NTA1",
"avatar_url": "https://avatars.githubusercontent.com/u/72618505?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jinlHe",
"html_url": "https://github.com/jinlHe",
"followers_url": "https://api.github.com/users/jinlHe/fo... | [] | closed | false | null | [] | null | 4 | 2024-04-01T15:06:12 | 2024-06-08T15:03:50 | 2024-06-08T15:03:50 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I got exactly the same error as the author @mahdip72 .
> RuntimeError: Expected to have finished reduction in the prior iteration before starting a new one. This error indicates that your module has parameters that were not used in producing loss. You can enable unused parameter detection by passing the keyword argum... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1612/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1612/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1611 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1611/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1611/comments | https://api.github.com/repos/huggingface/peft/issues/1611/events | https://github.com/huggingface/peft/issues/1611 | 2,218,333,955 | I_kwDOIf9iDM6EORcD | 1,611 | More Input: If I use PyTorch Lightning for large model fine-tuning (LoRA) tasks, and I want to pass an additional parameter p outside the input-output pair (input, label), to be used in the Linear layer of LoRA for some special tasks during forward pass, how should I proceed? | {
"login": "AkaliKong",
"id": 44406991,
"node_id": "MDQ6VXNlcjQ0NDA2OTkx",
"avatar_url": "https://avatars.githubusercontent.com/u/44406991?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AkaliKong",
"html_url": "https://github.com/AkaliKong",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 2 | 2024-04-01T13:58:55 | 2024-05-09T15:04:16 | 2024-05-09T15:04:16 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | How can I make it?
If I use PyTorch Lightning for large model fine-tuning (LoRA) tasks, and I want to pass an additional parameter p outside the input-output pair (input, label), to be used in the Linear layer of LoRA for some special tasks during forward pass, how should I proceed? | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1611/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1611/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1609 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1609/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1609/comments | https://api.github.com/repos/huggingface/peft/issues/1609/events | https://github.com/huggingface/peft/issues/1609 | 2,217,740,169 | I_kwDOIf9iDM6EMAeJ | 1,609 | AttributeError: ModulesToSaveWrapper has no attribute layers | {
"login": "pxy963",
"id": 72979645,
"node_id": "MDQ6VXNlcjcyOTc5NjQ1",
"avatar_url": "https://avatars.githubusercontent.com/u/72979645?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pxy963",
"html_url": "https://github.com/pxy963",
"followers_url": "https://api.github.com/users/pxy963/fo... | [] | closed | false | null | [] | null | 4 | 2024-04-01T07:45:06 | 2024-06-14T09:33:09 | 2024-04-03T03:40:49 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
accelerate 0.28.0
peft 0.10.0
torch == 2.2.1
transformer==4.39.2
flash-attn==2.5.6
python == 3.10
### Who can help?
@BenjaminBossan
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported ta... | {
"login": "pxy963",
"id": 72979645,
"node_id": "MDQ6VXNlcjcyOTc5NjQ1",
"avatar_url": "https://avatars.githubusercontent.com/u/72979645?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pxy963",
"html_url": "https://github.com/pxy963",
"followers_url": "https://api.github.com/users/pxy963/fo... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1609/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1609/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1610 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1610/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1610/comments | https://api.github.com/repos/huggingface/peft/issues/1610/events | https://github.com/huggingface/peft/issues/1610 | 2,218,037,357 | I_kwDOIf9iDM6ENJBt | 1,610 | Prefix tuning configuration issue | {
"login": "AswiniNLP",
"id": 132350955,
"node_id": "U_kgDOB-OD6w",
"avatar_url": "https://avatars.githubusercontent.com/u/132350955?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AswiniNLP",
"html_url": "https://github.com/AswiniNLP",
"followers_url": "https://api.github.com/users/Aswini... | [] | closed | false | null | [] | null | 1 | 2024-03-31T09:31:51 | 2024-04-02T12:54:22 | 2024-04-02T12:54:22 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
When ever I am using peft_config = PrefixTuningConfig(
peft_type="PREFIX_TUNING",
inference_mode=False,
task_type=TaskType.SEQ_2_SEQ_LM,
num_virtual_tokens=20,
token_dim=768,
num_transformer_submodules=1,
num_attention_heads=12,
num_layers=12,
encoder_hidden... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1610/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1610/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1608 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1608/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1608/comments | https://api.github.com/repos/huggingface/peft/issues/1608/events | https://github.com/huggingface/peft/issues/1608 | 2,216,867,734 | I_kwDOIf9iDM6EIreW | 1,608 | Configuration issue | {
"login": "AswiniNLP",
"id": 132350955,
"node_id": "U_kgDOB-OD6w",
"avatar_url": "https://avatars.githubusercontent.com/u/132350955?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AswiniNLP",
"html_url": "https://github.com/AswiniNLP",
"followers_url": "https://api.github.com/users/Aswini... | [] | closed | false | null | [] | null | 11 | 2024-03-31T09:28:51 | 2024-04-08T09:51:35 | 2024-04-05T08:48:09 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
When ever I am using
```python
peft_config = PrefixTuningConfig(
peft_type="PREFIX_TUNING",
inference_mode=False,
task_type=TaskType.SEQ_2_SEQ_LM,
num_virtual_tokens=20,
token_dim=768,
num_transformer_submodules=1,
num_attention_heads=12,
num_layers=12... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1608/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1608/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1607 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1607/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1607/comments | https://api.github.com/repos/huggingface/peft/issues/1607/events | https://github.com/huggingface/peft/issues/1607 | 2,216,824,348 | I_kwDOIf9iDM6EIg4c | 1,607 | p_tuning layers disappear after loading peft model | {
"login": "victoriousMan",
"id": 19256639,
"node_id": "MDQ6VXNlcjE5MjU2NjM5",
"avatar_url": "https://avatars.githubusercontent.com/u/19256639?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/victoriousMan",
"html_url": "https://github.com/victoriousMan",
"followers_url": "https://api.githu... | [] | closed | false | null | [] | null | 4 | 2024-03-31T07:29:35 | 2024-04-05T01:36:41 | 2024-04-05T01:36:17 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | After I trained my peft model (in this case P_Tuning), I save it using save_pretrained method. Later I load the model using:
```
config.peft_config = PeftConfig.from_pretrained(peft_model_id,
inference_mode=inference)
MODEL_TASK_TYPE = {
TaskType.CAUSAL_LM: AutoMode... | {
"login": "victoriousMan",
"id": 19256639,
"node_id": "MDQ6VXNlcjE5MjU2NjM5",
"avatar_url": "https://avatars.githubusercontent.com/u/19256639?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/victoriousMan",
"html_url": "https://github.com/victoriousMan",
"followers_url": "https://api.githu... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1607/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1607/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1606 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1606/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1606/comments | https://api.github.com/repos/huggingface/peft/issues/1606/events | https://github.com/huggingface/peft/issues/1606 | 2,216,061,963 | I_kwDOIf9iDM6EFmwL | 1,606 | Using LoRA on custom models | {
"login": "wsuSaiman",
"id": 156061192,
"node_id": "U_kgDOCU1OCA",
"avatar_url": "https://avatars.githubusercontent.com/u/156061192?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/wsuSaiman",
"html_url": "https://github.com/wsuSaiman",
"followers_url": "https://api.github.com/users/wsuSai... | [] | closed | false | null | [] | null | 2 | 2024-03-29T22:42:56 | 2024-05-08T15:03:32 | 2024-05-08T15:03:32 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
I am looking to train a full MLP model and train the model using LoRA later. For few epochs the MLP is run and it is assumed as pretrained, then we apply LoRA. How can this implementation be possible?
### Motivation
Related to implementation of LoRA
### Your contribution
I am trying to use this... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1606/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1606/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1605 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1605/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1605/comments | https://api.github.com/repos/huggingface/peft/issues/1605/events | https://github.com/huggingface/peft/issues/1605 | 2,214,776,979 | I_kwDOIf9iDM6EAtCT | 1,605 | Is there a choice to use `ignore_mismatch_sizes` in PeftModel.from_pretrained like AutoModel in Transformers? | {
"login": "haotong-yang",
"id": 56245046,
"node_id": "MDQ6VXNlcjU2MjQ1MDQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/56245046?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/haotong-yang",
"html_url": "https://github.com/haotong-yang",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | null | 7 | 2024-03-29T06:49:58 | 2024-04-08T10:23:36 | 2024-04-08T10:23:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Add ignore_mismatch_sizes in PeftModel.from_pretrained.
### Motivation
When I try to add some special tokens into an LM and use an adapter, I find though I can set ignore_mismatch_sizes = True when I load the base model, I have no similar choice to ignore the mismatch parameters when I load the a... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1605/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1605/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1604 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1604/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1604/comments | https://api.github.com/repos/huggingface/peft/issues/1604/events | https://github.com/huggingface/peft/issues/1604 | 2,214,743,587 | I_kwDOIf9iDM6EAk4j | 1,604 | [Feature Request] Support `torch.nn.Conv1d` layers for LoRA | {
"login": "dg845",
"id": 58458699,
"node_id": "MDQ6VXNlcjU4NDU4Njk5",
"avatar_url": "https://avatars.githubusercontent.com/u/58458699?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dg845",
"html_url": "https://github.com/dg845",
"followers_url": "https://api.github.com/users/dg845/follow... | [] | closed | false | null | [] | null | 2 | 2024-03-29T06:25:11 | 2024-05-07T15:03:25 | 2024-05-07T15:03:25 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
I would like [`torch.nn.Conv1d`](https://pytorch.org/docs/stable/generated/torch.nn.Conv1d.html) layers to be supported by LoRA. Currently, if I try to include a `torch.nn.Conv1d` layer in `LoraConfig.target_modules`, I get the following error:
```bash
ValueError: Target module Conv1d(...) is... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1604/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1604/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1601 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1601/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1601/comments | https://api.github.com/repos/huggingface/peft/issues/1601/events | https://github.com/huggingface/peft/issues/1601 | 2,213,253,630 | I_kwDOIf9iDM6D65H- | 1,601 | LISA | {
"login": "timothelaborie",
"id": 97834767,
"node_id": "U_kgDOBdTXDw",
"avatar_url": "https://avatars.githubusercontent.com/u/97834767?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/timothelaborie",
"html_url": "https://github.com/timothelaborie",
"followers_url": "https://api.github.com... | [] | closed | false | null | [] | null | 7 | 2024-03-28T13:31:01 | 2024-06-02T15:03:43 | 2024-06-02T15:03:43 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Suggestion: Add a "use_lisa" parameter to use [this](https://arxiv.org/abs/2403.17919)
### Motivation
It's better than LoRA
### Your contribution
There is an implementation [here](https://github.com/OptimalScale/LMFlow/blob/be2c49c6a17ba43c6c9010a85cc63840102463c2/src/lmflow/pipeline/finetuner.... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1601/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 1
} | https://api.github.com/repos/huggingface/peft/issues/1601/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1599 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1599/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1599/comments | https://api.github.com/repos/huggingface/peft/issues/1599/events | https://github.com/huggingface/peft/issues/1599 | 2,212,776,224 | I_kwDOIf9iDM6D5Ekg | 1,599 | Error while merge and unload: InvalidHeaderDeserialization when open .safetensor model | {
"login": "dipeshpaulsystango",
"id": 154497724,
"node_id": "U_kgDOCTVyvA",
"avatar_url": "https://avatars.githubusercontent.com/u/154497724?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dipeshpaulsystango",
"html_url": "https://github.com/dipeshpaulsystango",
"followers_url": "https://... | [] | closed | false | null | [] | null | 12 | 2024-03-28T09:44:09 | 2024-03-28T17:58:17 | 2024-03-28T13:39:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
tensorflow==2.16.1
pip install git+https://github.com/huggingface/transformers
pip install git+https://github.com/huggingface/accelerate
pip install git+https://github.com/huggingface/peft
pip install git+https://github.com/huggingface/trl
pip install git+https://github.com/huggingface/datatrove... | {
"login": "dipeshpaulsystango",
"id": 154497724,
"node_id": "U_kgDOCTVyvA",
"avatar_url": "https://avatars.githubusercontent.com/u/154497724?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dipeshpaulsystango",
"html_url": "https://github.com/dipeshpaulsystango",
"followers_url": "https://... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1599/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1599/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1598 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1598/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1598/comments | https://api.github.com/repos/huggingface/peft/issues/1598/events | https://github.com/huggingface/peft/issues/1598 | 2,212,305,401 | I_kwDOIf9iDM6D3Rn5 | 1,598 | examples/sft/run_peft.sh model load dtype error | {
"login": "galtay",
"id": 663051,
"node_id": "MDQ6VXNlcjY2MzA1MQ==",
"avatar_url": "https://avatars.githubusercontent.com/u/663051?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/galtay",
"html_url": "https://github.com/galtay",
"followers_url": "https://api.github.com/users/galtay/follow... | [] | closed | false | null | [] | null | 7 | 2024-03-28T04:05:47 | 2024-04-04T18:44:58 | 2024-04-04T18:44:58 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft installed from main
c4c826c0c0dfe06620db8b8adcfa55f22acfed61
- `transformers` version: 4.39.1
- Platform: Linux-6.5.0-26-generic-x86_64-with-glibc2.35
- Python version: 3.11.8
- Huggingface_hub version: 0.22.0
- Safetensors version: 0.4.2
- Accelerate version: 0.28.0
- PyTorch version ... | {
"login": "pacman100",
"id": 13534540,
"node_id": "MDQ6VXNlcjEzNTM0NTQw",
"avatar_url": "https://avatars.githubusercontent.com/u/13534540?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pacman100",
"html_url": "https://github.com/pacman100",
"followers_url": "https://api.github.com/users/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1598/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1598/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1597 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1597/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1597/comments | https://api.github.com/repos/huggingface/peft/issues/1597/events | https://github.com/huggingface/peft/issues/1597 | 2,211,461,726 | I_kwDOIf9iDM6D0Dpe | 1,597 | RuntimeError: "addmm_impl_cpu_" not implemented for 'Half | {
"login": "azuryl",
"id": 23340671,
"node_id": "MDQ6VXNlcjIzMzQwNjcx",
"avatar_url": "https://avatars.githubusercontent.com/u/23340671?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/azuryl",
"html_url": "https://github.com/azuryl",
"followers_url": "https://api.github.com/users/azuryl/fo... | [] | closed | false | null | [] | null | 11 | 2024-03-27T18:05:30 | 2024-04-29T09:36:36 | 2024-04-29T09:36:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft v0.10
### Who can help?
@pacman100 @younesbelkada @sayakpaul
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reprodu... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1597/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1597/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1596 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1596/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1596/comments | https://api.github.com/repos/huggingface/peft/issues/1596/events | https://github.com/huggingface/peft/issues/1596 | 2,210,988,056 | I_kwDOIf9iDM6DyQAY | 1,596 | TypeError: ChatGLMForConditionalGeneration.forward() got an unexpected keyword argument 'decoder_input_ids' | {
"login": "12915494174",
"id": 68487349,
"node_id": "MDQ6VXNlcjY4NDg3MzQ5",
"avatar_url": "https://avatars.githubusercontent.com/u/68487349?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/12915494174",
"html_url": "https://github.com/12915494174",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | null | 4 | 2024-03-27T14:39:54 | 2024-03-28T10:09:09 | 2024-03-28T10:09:09 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
python 3.10.8
peft 0.7.1
transformers 4.38.1
datasets 2.18.0
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
... | {
"login": "12915494174",
"id": 68487349,
"node_id": "MDQ6VXNlcjY4NDg3MzQ5",
"avatar_url": "https://avatars.githubusercontent.com/u/68487349?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/12915494174",
"html_url": "https://github.com/12915494174",
"followers_url": "https://api.github.com/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1596/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1596/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1595 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1595/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1595/comments | https://api.github.com/repos/huggingface/peft/issues/1595/events | https://github.com/huggingface/peft/issues/1595 | 2,210,328,306 | I_kwDOIf9iDM6Dvu7y | 1,595 | Merging models and feature extraction | {
"login": "tyemelya",
"id": 117215512,
"node_id": "U_kgDOBvyRGA",
"avatar_url": "https://avatars.githubusercontent.com/u/117215512?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tyemelya",
"html_url": "https://github.com/tyemelya",
"followers_url": "https://api.github.com/users/tyemelya/... | [] | closed | false | null | [] | null | 4 | 2024-03-27T09:48:21 | 2024-03-27T15:01:03 | 2024-03-27T15:01:03 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello.
I used an example for fine-tuning with the LoRA technique for the image classification task. I fine-tuned with LoRA one transformer model for 2 different image classification tasks (with the same parameters and dataset, the difference is only in classes). Then, I want to merge the base model with received adapt... | {
"login": "tyemelya",
"id": 117215512,
"node_id": "U_kgDOBvyRGA",
"avatar_url": "https://avatars.githubusercontent.com/u/117215512?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tyemelya",
"html_url": "https://github.com/tyemelya",
"followers_url": "https://api.github.com/users/tyemelya/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1595/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1595/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1594 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1594/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1594/comments | https://api.github.com/repos/huggingface/peft/issues/1594/events | https://github.com/huggingface/peft/issues/1594 | 2,210,287,254 | I_kwDOIf9iDM6Dvk6W | 1,594 | model.print_trainable_parameters() is incomplete. | {
"login": "Minami-su",
"id": 76865636,
"node_id": "MDQ6VXNlcjc2ODY1NjM2",
"avatar_url": "https://avatars.githubusercontent.com/u/76865636?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Minami-su",
"html_url": "https://github.com/Minami-su",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 5 | 2024-03-27T09:28:06 | 2024-05-05T15:03:38 | 2024-05-05T15:03:38 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | model:ISTA-DASLab/Mixtral-8x7B-Instruct-v0_1-AQLM-2Bit-1x16-hf
peft:0.10.0
```
model = AutoModelForCausalLM.from_pretrained(base_model,
trust_remote_code=True,
attn_implementation="flash_attention_2",
... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1594/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1594/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1593 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1593/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1593/comments | https://api.github.com/repos/huggingface/peft/issues/1593/events | https://github.com/huggingface/peft/issues/1593 | 2,209,950,468 | I_kwDOIf9iDM6DuSsE | 1,593 | Getting Dora Model Is Very Slow | {
"login": "mallorbc",
"id": 39721523,
"node_id": "MDQ6VXNlcjM5NzIxNTIz",
"avatar_url": "https://avatars.githubusercontent.com/u/39721523?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mallorbc",
"html_url": "https://github.com/mallorbc",
"followers_url": "https://api.github.com/users/mal... | [] | closed | false | null | [] | null | 10 | 2024-03-27T06:05:56 | 2024-05-16T21:18:34 | 2024-05-05T15:03:39 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
Package Version
------------------------ ---------------
accelerate 0.29.0.dev0
aiohttp 3.9.3
aiosignal 1.3.1
annotated-types 0.6.0
appdirs 1.4.4
async-timeout 4.0.3
attrs 23.2... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1593/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1593/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1592 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1592/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1592/comments | https://api.github.com/repos/huggingface/peft/issues/1592/events | https://github.com/huggingface/peft/issues/1592 | 2,209,745,962 | I_kwDOIf9iDM6Dtgwq | 1,592 | ... | {
"login": "Minami-su",
"id": 76865636,
"node_id": "MDQ6VXNlcjc2ODY1NjM2",
"avatar_url": "https://avatars.githubusercontent.com/u/76865636?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Minami-su",
"html_url": "https://github.com/Minami-su",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 0 | 2024-03-27T03:31:09 | 2024-03-27T03:31:31 | 2024-03-27T03:31:24 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | .... | {
"login": "Minami-su",
"id": 76865636,
"node_id": "MDQ6VXNlcjc2ODY1NjM2",
"avatar_url": "https://avatars.githubusercontent.com/u/76865636?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Minami-su",
"html_url": "https://github.com/Minami-su",
"followers_url": "https://api.github.com/users/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1592/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1592/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1590 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1590/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1590/comments | https://api.github.com/repos/huggingface/peft/issues/1590/events | https://github.com/huggingface/peft/issues/1590 | 2,208,399,608 | I_kwDOIf9iDM6DoYD4 | 1,590 | NotImplementedError: Saving is not supported for PeftMixedModel (yet). | {
"login": "mickeymickeymonkey",
"id": 142171154,
"node_id": "U_kgDOCHlcEg",
"avatar_url": "https://avatars.githubusercontent.com/u/142171154?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mickeymickeymonkey",
"html_url": "https://github.com/mickeymickeymonkey",
"followers_url": "https://... | [] | closed | false | null | [] | null | 2 | 2024-03-26T14:20:47 | 2024-03-28T12:08:30 | 2024-03-28T12:08:29 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
I trained a peftmixedmodel with different adapters but can't save the adapter weights. any idea what why it is not possible to save_pretrained like normal? any idea when this can be possible?
### Motivation
being able to save the adapter weights only is a massive save on storage
### Your contrib... | {
"login": "mickeymickeymonkey",
"id": 142171154,
"node_id": "U_kgDOCHlcEg",
"avatar_url": "https://avatars.githubusercontent.com/u/142171154?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mickeymickeymonkey",
"html_url": "https://github.com/mickeymickeymonkey",
"followers_url": "https://... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1590/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1590/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1589 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1589/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1589/comments | https://api.github.com/repos/huggingface/peft/issues/1589/events | https://github.com/huggingface/peft/issues/1589 | 2,206,971,858 | I_kwDOIf9iDM6Di7fS | 1,589 | TaskType attribute in Lora config for question answering is wrong | {
"login": "aryangoyal7",
"id": 98456095,
"node_id": "U_kgDOBd5SHw",
"avatar_url": "https://avatars.githubusercontent.com/u/98456095?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/aryangoyal7",
"html_url": "https://github.com/aryangoyal7",
"followers_url": "https://api.github.com/users/ar... | [] | closed | false | null | [] | null | 4 | 2024-03-26T00:19:15 | 2024-05-05T15:03:41 | 2024-05-05T15:03:41 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
in documentation it is "QUESTION_ANS = "QUESTION_ANS"" but it doesnt work, i test the same code for CASUAL_LM , it works, hence needs to be fixed, I tried "QUESTION_ANSWERING" that does work as wellthe
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- ... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1589/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1589/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1586 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1586/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1586/comments | https://api.github.com/repos/huggingface/peft/issues/1586/events | https://github.com/huggingface/peft/issues/1586 | 2,205,665,424 | I_kwDOIf9iDM6Dd8iQ | 1,586 | Error in LoraModel docstring | {
"login": "brynhayder",
"id": 4146799,
"node_id": "MDQ6VXNlcjQxNDY3OTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/4146799?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/brynhayder",
"html_url": "https://github.com/brynhayder",
"followers_url": "https://api.github.com/users... | [] | closed | false | null | [] | null | 3 | 2024-03-25T12:43:25 | 2024-05-03T15:03:30 | 2024-05-03T15:03:30 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | https://github.com/huggingface/peft/blob/78daa4cf7624ec9c2c38fe73b0317b61428ca26b/src/peft/tuners/lora/model.py#L71
Not possible to default because written as a positional argument | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1586/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1586/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1583 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1583/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1583/comments | https://api.github.com/repos/huggingface/peft/issues/1583/events | https://github.com/huggingface/peft/issues/1583 | 2,204,228,467 | I_kwDOIf9iDM6DYdtz | 1,583 | cannot import name 'prepare_model_for_int8_training' from 'peft' | {
"login": "Eyict",
"id": 73412378,
"node_id": "MDQ6VXNlcjczNDEyMzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/73412378?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Eyict",
"html_url": "https://github.com/Eyict",
"followers_url": "https://api.github.com/users/Eyict/follow... | [] | closed | false | null | [] | null | 8 | 2024-03-24T07:44:42 | 2024-05-15T13:49:13 | 2024-04-25T09:24:37 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | When I run the finetuning sample code from llama-recipes i.e. peft_finetuning.ipynb, I get this error. My code is , and python=3.9, peft=0.10.0
the error is
![image](https://github.com/huggingface/peft/assets/73412378... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1583/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1583/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1580 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1580/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1580/comments | https://api.github.com/repos/huggingface/peft/issues/1580/events | https://github.com/huggingface/peft/issues/1580 | 2,200,112,393 | I_kwDOIf9iDM6DIw0J | 1,580 | PeftModel is_trainable=True causes generate output to be garbage. | {
"login": "o1lo01ol1o",
"id": 8008053,
"node_id": "MDQ6VXNlcjgwMDgwNTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/8008053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/o1lo01ol1o",
"html_url": "https://github.com/o1lo01ol1o",
"followers_url": "https://api.github.com/users... | [] | closed | false | null | [] | null | 6 | 2024-03-21T12:49:53 | 2024-05-08T15:03:34 | 2024-05-08T15:03:34 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null |
### TL:DR
When loading a pretrained base model and then subsequently loading a trained PeftModel adaptor on `mistralai/Mistral-7B-Instruct-v0.2`, `model.generate()` behaves as expected. However, if the `PeftModel` has is_trainable set to `True`, the output is garbage.
Example:
```python
base_model = "m... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1580/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1580/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1579 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1579/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1579/comments | https://api.github.com/repos/huggingface/peft/issues/1579/events | https://github.com/huggingface/peft/issues/1579 | 2,199,991,077 | I_kwDOIf9iDM6DITMl | 1,579 | error merge_and_unload for adapter with a prefix | {
"login": "afalf",
"id": 101537385,
"node_id": "U_kgDOBg1WaQ",
"avatar_url": "https://avatars.githubusercontent.com/u/101537385?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/afalf",
"html_url": "https://github.com/afalf",
"followers_url": "https://api.github.com/users/afalf/followers",
... | [] | closed | false | null | [] | null | 23 | 2024-03-21T11:58:38 | 2024-04-29T15:03:46 | 2024-04-29T15:03:46 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft version: 0.9.0
transforemrs version: 4.37.2
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [X] My own task or dataset (give details below)
### Reproduct... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1579/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1579/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1577 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1577/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1577/comments | https://api.github.com/repos/huggingface/peft/issues/1577/events | https://github.com/huggingface/peft/issues/1577 | 2,199,552,307 | I_kwDOIf9iDM6DGoEz | 1,577 | element 0 of tensors does not require grad and does not have a grad_fn | {
"login": "ghost",
"id": 10137,
"node_id": "MDQ6VXNlcjEwMTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/10137?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ghost",
"html_url": "https://github.com/ghost",
"followers_url": "https://api.github.com/users/ghost/followers",
"f... | [] | closed | false | null | [] | null | 4 | 2024-03-21T08:45:33 | 2024-04-07T01:13:11 | 2024-04-07T01:13:11 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
-
### Who can help?
@pacman100 @younesbelkada @sayakpaul
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
... | null | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1577/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1577/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1576 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1576/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1576/comments | https://api.github.com/repos/huggingface/peft/issues/1576/events | https://github.com/huggingface/peft/issues/1576 | 2,199,162,595 | I_kwDOIf9iDM6DFI7j | 1,576 | Loading LORA weights in `diffusers` with a `peft` backend increases in latency as more paths are added to `PYTHONPATH` | {
"login": "tisles",
"id": 54301847,
"node_id": "MDQ6VXNlcjU0MzAxODQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/54301847?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tisles",
"html_url": "https://github.com/tisles",
"followers_url": "https://api.github.com/users/tisles/fo... | [] | closed | false | null | [] | null | 4 | 2024-03-21T05:36:47 | 2024-03-26T13:28:23 | 2024-03-26T13:28:23 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```
accelerate==0.21.0
diffusers==0.26.3
peft==0.9.0
safetensors==0.3.3
tokenizers==0.15.2
torch==2.2.1
transformers==4.36.2
```
### Who can help?
@sayakpaul
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supp... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1576/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1576/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1575 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1575/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1575/comments | https://api.github.com/repos/huggingface/peft/issues/1575/events | https://github.com/huggingface/peft/issues/1575 | 2,198,754,941 | I_kwDOIf9iDM6DDlZ9 | 1,575 | MPS: Cannot add LoRA to Unet (LoftQ) | {
"login": "bghira",
"id": 59658056,
"node_id": "MDQ6VXNlcjU5NjU4MDU2",
"avatar_url": "https://avatars.githubusercontent.com/u/59658056?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bghira",
"html_url": "https://github.com/bghira",
"followers_url": "https://api.github.com/users/bghira/fo... | [] | closed | false | null | [] | null | 6 | 2024-03-20T23:09:13 | 2024-04-05T12:35:33 | 2024-04-05T12:35:33 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft v0.9.0
accelerate v0.26.1
apple m3 max
### Who can help?
@sayakpaul
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
... | {
"login": "bghira",
"id": 59658056,
"node_id": "MDQ6VXNlcjU5NjU4MDU2",
"avatar_url": "https://avatars.githubusercontent.com/u/59658056?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bghira",
"html_url": "https://github.com/bghira",
"followers_url": "https://api.github.com/users/bghira/fo... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1575/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1575/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1574 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1574/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1574/comments | https://api.github.com/repos/huggingface/peft/issues/1574/events | https://github.com/huggingface/peft/issues/1574 | 2,198,230,904 | I_kwDOIf9iDM6DBld4 | 1,574 | 'set_adapter()' throws "ValueError: Adapter not found in odict_keys" after 'load_adapter()' | {
"login": "YnezT0311",
"id": 59566630,
"node_id": "MDQ6VXNlcjU5NTY2NjMw",
"avatar_url": "https://avatars.githubusercontent.com/u/59566630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YnezT0311",
"html_url": "https://github.com/YnezT0311",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 11 | 2024-03-20T18:12:04 | 2024-04-09T10:59:26 | 2024-04-09T10:59:26 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft-0.9.0
### Who can help?
@pacman100
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
Saved two finetune... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1574/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1574/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1572 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1572/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1572/comments | https://api.github.com/repos/huggingface/peft/issues/1572/events | https://github.com/huggingface/peft/issues/1572 | 2,196,936,296 | I_kwDOIf9iDM6C8pZo | 1,572 | PP or TP supported for multi-node training? | {
"login": "ghost",
"id": 10137,
"node_id": "MDQ6VXNlcjEwMTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/10137?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ghost",
"html_url": "https://github.com/ghost",
"followers_url": "https://api.github.com/users/ghost/followers",
"f... | [] | closed | false | null | [] | null | 3 | 2024-03-20T08:25:41 | 2024-03-21T08:34:56 | 2024-03-21T08:34:56 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
deepspeed
accelerate
### Who can help?
@pacman100 @younesbelkada @sayakpaul
### Information
- [ ] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproductio... | null | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1572/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1572/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1571 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1571/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1571/comments | https://api.github.com/repos/huggingface/peft/issues/1571/events | https://github.com/huggingface/peft/issues/1571 | 2,196,786,021 | I_kwDOIf9iDM6C8Etl | 1,571 | Custom Training for multiple LoRAs for the same model. | {
"login": "bellos1203",
"id": 28821630,
"node_id": "MDQ6VXNlcjI4ODIxNjMw",
"avatar_url": "https://avatars.githubusercontent.com/u/28821630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bellos1203",
"html_url": "https://github.com/bellos1203",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | null | 4 | 2024-03-20T06:52:33 | 2024-03-21T08:30:29 | 2024-03-21T08:30:29 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi, I'm trying to train multiple LoRA modules for my model.
Here is my situation.
I have a large model consisting of two modules, A and B.
For input data, I have x and y, which have different characteristics from each other.
What I want is two LoRA modules for A, which are A_x, and A_y.
Then, the output of A_x a... | {
"login": "bellos1203",
"id": 28821630,
"node_id": "MDQ6VXNlcjI4ODIxNjMw",
"avatar_url": "https://avatars.githubusercontent.com/u/28821630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bellos1203",
"html_url": "https://github.com/bellos1203",
"followers_url": "https://api.github.com/use... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1571/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1571/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1569 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1569/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1569/comments | https://api.github.com/repos/huggingface/peft/issues/1569/events | https://github.com/huggingface/peft/issues/1569 | 2,193,554,005 | I_kwDOIf9iDM6CvvpV | 1,569 | Error saving LoRaModel with Wav2vec2forCTC basemodel | {
"login": "geoffvdr",
"id": 47786797,
"node_id": "MDQ6VXNlcjQ3Nzg2Nzk3",
"avatar_url": "https://avatars.githubusercontent.com/u/47786797?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/geoffvdr",
"html_url": "https://github.com/geoffvdr",
"followers_url": "https://api.github.com/users/geo... | [] | closed | false | null | [] | null | 7 | 2024-03-18T22:37:58 | 2024-05-13T15:04:20 | 2024-05-13T15:04:20 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello,
I can successfully train a LoraModel with Wav2vec2forCTC as basemodel, however the checkpoint saving fails because the basemodel does not implement `get_input_embeddings()`.
Error traceback:
```
Traceback (most recent call last):
2024-03-18T17:21:06.765068910+01:00 File "/project_antwerp/huggingface/t... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1569/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1569/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1568 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1568/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1568/comments | https://api.github.com/repos/huggingface/peft/issues/1568/events | https://github.com/huggingface/peft/issues/1568 | 2,192,350,021 | I_kwDOIf9iDM6CrJtF | 1,568 | When peft>=0.7.0, fine-tuning ChatGLM3-6B causes the model to become dumb with a loss of 0 | {
"login": "Tangent-90C",
"id": 28804414,
"node_id": "MDQ6VXNlcjI4ODA0NDE0",
"avatar_url": "https://avatars.githubusercontent.com/u/28804414?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Tangent-90C",
"html_url": "https://github.com/Tangent-90C",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | null | 13 | 2024-03-18T14:17:00 | 2024-04-11T15:22:16 | 2024-04-10T15:25:53 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
```
Package Version
------------------------------ ---------------
accelerate 0.28.0
datasets 2.18.0
peft 0.9.0
torch 2.1.2+cu121
transformers 4.38.2
```
### Wh... | {
"login": "BenjaminBossan",
"id": 6229650,
"node_id": "MDQ6VXNlcjYyMjk2NTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/6229650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BenjaminBossan",
"html_url": "https://github.com/BenjaminBossan",
"followers_url": "https://api.gith... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1568/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1568/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1567 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1567/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1567/comments | https://api.github.com/repos/huggingface/peft/issues/1567/events | https://github.com/huggingface/peft/issues/1567 | 2,190,221,591 | I_kwDOIf9iDM6CjCEX | 1,567 | Base Model Revision | {
"login": "mnoukhov",
"id": 3391297,
"node_id": "MDQ6VXNlcjMzOTEyOTc=",
"avatar_url": "https://avatars.githubusercontent.com/u/3391297?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mnoukhov",
"html_url": "https://github.com/mnoukhov",
"followers_url": "https://api.github.com/users/mnouk... | [] | closed | false | null | [] | null | 8 | 2024-03-16T19:52:36 | 2024-06-06T15:05:11 | 2024-06-06T15:05:10 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
Currently, the `adapter_config` only stores the `base_model_name_or_path` but not the `revision`. This means that if a `PeftModel` is trained on top of a model with a `revision` then loading `PeftModel.from_pretrained` will load the `main` revision of the base model, not the correct revision.
W... | {
"login": "mnoukhov",
"id": 3391297,
"node_id": "MDQ6VXNlcjMzOTEyOTc=",
"avatar_url": "https://avatars.githubusercontent.com/u/3391297?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mnoukhov",
"html_url": "https://github.com/mnoukhov",
"followers_url": "https://api.github.com/users/mnouk... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1567/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1567/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1566 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1566/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1566/comments | https://api.github.com/repos/huggingface/peft/issues/1566/events | https://github.com/huggingface/peft/issues/1566 | 2,188,250,336 | I_kwDOIf9iDM6Cbgzg | 1,566 | cannot load int8/4 model with deepspeed zero3 | {
"login": "ghost",
"id": 10137,
"node_id": "MDQ6VXNlcjEwMTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/10137?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ghost",
"html_url": "https://github.com/ghost",
"followers_url": "https://api.github.com/users/ghost/followers",
"f... | [] | closed | false | null | [] | null | 5 | 2024-03-15T10:59:41 | 2024-03-18T02:51:11 | 2024-03-18T02:51:11 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
--
### Who can help?
@pacman100 @younesbelkada
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
Hello, when I tried to run th... | null | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1566/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1566/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1563 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1563/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1563/comments | https://api.github.com/repos/huggingface/peft/issues/1563/events | https://github.com/huggingface/peft/issues/1563 | 2,186,446,802 | I_kwDOIf9iDM6CUofS | 1,563 | Add a new (not so new and it is typical) fine-tuning method called VPT | {
"login": "2catycm",
"id": 41530341,
"node_id": "MDQ6VXNlcjQxNTMwMzQx",
"avatar_url": "https://avatars.githubusercontent.com/u/41530341?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/2catycm",
"html_url": "https://github.com/2catycm",
"followers_url": "https://api.github.com/users/2catyc... | [] | closed | false | null | [] | null | 3 | 2024-03-14T13:53:38 | 2024-04-21T15:03:35 | 2024-04-21T15:03:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
1. link to the source.
https://arxiv.org/abs/2203.12119
### Motivation
2. evidence there is general interest in using the method
- widely compared in image realted peft research.
### Your contribution
I am not sure whether P-Tuning is enough.
applying P-Tuning on ViT = VPT?
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1563/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1563/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1562 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1562/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1562/comments | https://api.github.com/repos/huggingface/peft/issues/1562/events | https://github.com/huggingface/peft/issues/1562 | 2,185,736,715 | I_kwDOIf9iDM6CR7IL | 1,562 | How do I easily inherit and register the new method? | {
"login": "mrwu-mac",
"id": 59433903,
"node_id": "MDQ6VXNlcjU5NDMzOTAz",
"avatar_url": "https://avatars.githubusercontent.com/u/59433903?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mrwu-mac",
"html_url": "https://github.com/mrwu-mac",
"followers_url": "https://api.github.com/users/mrw... | [] | closed | false | null | [] | null | 2 | 2024-03-14T08:26:53 | 2024-03-14T10:30:42 | 2024-03-14T10:30:42 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Is there an example of inheriting existing methods and registering my own new methods, I guess this will help enhance the extensibility of the repository. Such as, how can i add a linear layer for Lora? | {
"login": "mrwu-mac",
"id": 59433903,
"node_id": "MDQ6VXNlcjU5NDMzOTAz",
"avatar_url": "https://avatars.githubusercontent.com/u/59433903?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mrwu-mac",
"html_url": "https://github.com/mrwu-mac",
"followers_url": "https://api.github.com/users/mrw... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1562/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1562/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1560 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1560/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1560/comments | https://api.github.com/repos/huggingface/peft/issues/1560/events | https://github.com/huggingface/peft/issues/1560 | 2,185,278,207 | I_kwDOIf9iDM6CQLL_ | 1,560 | Add a new fine-tuning method called Conv-Lora | {
"login": "2catycm",
"id": 41530341,
"node_id": "MDQ6VXNlcjQxNTMwMzQx",
"avatar_url": "https://avatars.githubusercontent.com/u/41530341?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/2catycm",
"html_url": "https://github.com/2catycm",
"followers_url": "https://api.github.com/users/2catyc... | [] | closed | false | null | [] | null | 2 | 2024-03-14T02:56:55 | 2024-04-21T15:03:36 | 2024-04-21T15:03:36 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### Feature request
1. link to the source.
This is an ICLR 2024 paper, titiled "Convolution Meets LoRA: Parameter Efficient Finetuning for Segment Anything Model".
Here is the openreview link:
https://openreview.net/forum?id=ezscMer8L0
### Motivation
2. evidence there is general interest in using the method
- ... | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1560/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1560/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1554 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1554/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1554/comments | https://api.github.com/repos/huggingface/peft/issues/1554/events | https://github.com/huggingface/peft/issues/1554 | 2,181,071,548 | I_kwDOIf9iDM6CAIK8 | 1,554 | size mismatch of embedding layer without adding any token | {
"login": "rangehow",
"id": 88258534,
"node_id": "MDQ6VXNlcjg4MjU4NTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/88258534?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rangehow",
"html_url": "https://github.com/rangehow",
"followers_url": "https://api.github.com/users/ran... | [] | closed | false | null | [] | null | 1 | 2024-03-12T08:59:37 | 2024-03-12T09:09:03 | 2024-03-12T09:09:03 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
peft: 0.9.1.dev0
transformers: 4.39.0
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [X] An officially supported task in the `examples` folder
- [ ] My own task or dataset (give details below)
### Reproduction
LORA... | {
"login": "rangehow",
"id": 88258534,
"node_id": "MDQ6VXNlcjg4MjU4NTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/88258534?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rangehow",
"html_url": "https://github.com/rangehow",
"followers_url": "https://api.github.com/users/ran... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1554/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1554/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1553 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1553/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1553/comments | https://api.github.com/repos/huggingface/peft/issues/1553/events | https://github.com/huggingface/peft/issues/1553 | 2,180,912,151 | I_kwDOIf9iDM6B_hQX | 1,553 | adding lora adapter to embedding layer while using bitsandbytes and mixed precision training gives "RuntimeError: a leaf Variable that requires grad is being used in an in-place operation." | {
"login": "cai-rishabh",
"id": 142016092,
"node_id": "U_kgDOCHb-XA",
"avatar_url": "https://avatars.githubusercontent.com/u/142016092?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cai-rishabh",
"html_url": "https://github.com/cai-rishabh",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 7 | 2024-03-12T07:24:49 | 2024-04-25T10:19:52 | 2024-03-12T10:58:19 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
------- nvidia-smi output ----------
Tesla T4
NVIDIA-SMI 545.23.08
Driver Version: 545.23.08
CUDA Version: 12.3
-------- hostnamectl output --------
Ubuntu 20.04.6 LTS
Kernel: Linux 5.15.0-1048-aws
Architecture: x86-64
--------- pip freeze output ----------
accele... | {
"login": "cai-rishabh",
"id": 142016092,
"node_id": "U_kgDOCHb-XA",
"avatar_url": "https://avatars.githubusercontent.com/u/142016092?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cai-rishabh",
"html_url": "https://github.com/cai-rishabh",
"followers_url": "https://api.github.com/users/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1553/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1553/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1549 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1549/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1549/comments | https://api.github.com/repos/huggingface/peft/issues/1549/events | https://github.com/huggingface/peft/issues/1549 | 2,177,991,397 | I_kwDOIf9iDM6B0YLl | 1,549 | adapter_config.json is saved as "loftq_config": { "loftq_bits": "4bit", ... } | {
"login": "lottopotato",
"id": 31846796,
"node_id": "MDQ6VXNlcjMxODQ2Nzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31846796?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lottopotato",
"html_url": "https://github.com/lottopotato",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | null | 1 | 2024-03-11T00:29:09 | 2024-03-11T00:44:15 | 2024-03-11T00:35:08 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
transformers 4.38.1
peft 0.8.2
### Report
Due to
https://github.com/huggingface/peft/blob/7e84dec20b3106bdd0a90ba8e80187f0aec835b7/src/peft/tuners/lora/layer.py#L151-L161
and
https://github.com/huggingface/peft/blob/7e84dec20b3106bdd0a90ba8e80187f0aec835b7/src/... | {
"login": "lottopotato",
"id": 31846796,
"node_id": "MDQ6VXNlcjMxODQ2Nzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/31846796?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lottopotato",
"html_url": "https://github.com/lottopotato",
"followers_url": "https://api.github.com/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1549/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1549/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1545 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1545/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1545/comments | https://api.github.com/repos/huggingface/peft/issues/1545/events | https://github.com/huggingface/peft/issues/1545 | 2,175,879,730 | I_kwDOIf9iDM6BsUoy | 1,545 | How to use lora finetune moe model | {
"login": "Minami-su",
"id": 76865636,
"node_id": "MDQ6VXNlcjc2ODY1NjM2",
"avatar_url": "https://avatars.githubusercontent.com/u/76865636?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Minami-su",
"html_url": "https://github.com/Minami-su",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | null | 2 | 2024-03-08T11:45:09 | 2024-04-16T15:03:39 | 2024-04-16T15:03:38 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | null | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url"... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1545/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/peft/issues/1545/timeline | null | completed | false |
https://api.github.com/repos/huggingface/peft/issues/1544 | https://api.github.com/repos/huggingface/peft | https://api.github.com/repos/huggingface/peft/issues/1544/labels{/name} | https://api.github.com/repos/huggingface/peft/issues/1544/comments | https://api.github.com/repos/huggingface/peft/issues/1544/events | https://github.com/huggingface/peft/issues/1544 | 2,175,529,947 | I_kwDOIf9iDM6Bq_Pb | 1,544 | Fail to use multi-GPUs with peft model | {
"login": "theodpzz",
"id": 101199304,
"node_id": "U_kgDOBggtyA",
"avatar_url": "https://avatars.githubusercontent.com/u/101199304?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/theodpzz",
"html_url": "https://github.com/theodpzz",
"followers_url": "https://api.github.com/users/theodpzz/... | [] | closed | false | null | [] | null | 3 | 2024-03-08T08:24:46 | 2024-03-11T13:13:25 | 2024-03-11T13:11:14 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ### System Info
transformers version: 4.38.1
accelerate version: 0.26.1
peft version: 0.8.2
pytorch version: 2.1.2
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own modified scripts
### Tasks
- [ ] An officially supported task in the `examples` fold... | {
"login": "theodpzz",
"id": 101199304,
"node_id": "U_kgDOBggtyA",
"avatar_url": "https://avatars.githubusercontent.com/u/101199304?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/theodpzz",
"html_url": "https://github.com/theodpzz",
"followers_url": "https://api.github.com/users/theodpzz/... | {
"url": "https://api.github.com/repos/huggingface/peft/issues/1544/reactions",
"total_count": 2,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 2
} | https://api.github.com/repos/huggingface/peft/issues/1544/timeline | null | completed | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.