Hugging Face's logo

anupk
/
out

PEFT
Safetensors
Generated from Trainer
Model card Files
xet
 Community
anupk commited on
Commit
104b0b7
·
verified ·
1 Parent(s): be51957

anupk/askPauladapter

Browse files
Files changed (5) hide show
  1. README.md +23 -28
  2. adapter_config.json +6 -5
  3. adapter_model.safetensors +2 -2
  4. tokenizer_config.json +1 -0
  5. training_args.bin +2 -2
README.md CHANGED
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 4.3633
20
 
21
  ## Model description
22
 
@@ -41,44 +41,39 @@ The following hyperparameters were used during training:
41
  - seed: 42
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: constant
44
- - num_epochs: 25
45
  - mixed_precision_training: Native AMP
46
 
47
  ### Training results
48
 
49
  | Training Loss | Epoch | Step | Validation Loss |
50
  |:-------------:|:-----:|:----:|:---------------:|
51
- | 3.0161 | 1.0 | 326 | 2.6890 |
52
- | 1.3622 | 2.0 | 652 | 2.9713 |
53
- | 1.0421 | 3.0 | 978 | 3.1679 |
54
- | 0.7533 | 4.0 | 1304 | 3.5433 |
55
- | 0.5052 | 5.0 | 1630 | 3.5190 |
56
- | 0.4463 | 6.0 | 1956 | 3.8213 |
57
- | 0.2102 | 7.0 | 2282 | 3.8646 |
58
- | 0.3904 | 8.0 | 2608 | 3.9794 |
59
- | 0.1932 | 9.0 | 2934 | 3.9933 |
60
- | 0.3121 | 10.0 | 3260 | 4.2430 |
61
- | 0.1707 | 11.0 | 3586 | 4.3414 |
62
- | 0.3446 | 12.0 | 3912 | 5.0113 |
63
- | 0.1671 | 13.0 | 4238 | 4.5196 |
64
- | 0.1743 | 14.0 | 4564 | 4.4975 |
65
- | 0.4551 | 15.0 | 4890 | 4.2461 |
66
- | 0.1981 | 16.0 | 5216 | 4.9300 |
67
- | 0.2151 | 17.0 | 5542 | 4.8182 |
68
- | 0.1077 | 18.0 | 5868 | 4.6348 |
69
- | 0.2005 | 19.0 | 6194 | 5.1244 |
70
- | 0.1163 | 20.0 | 6520 | 4.6448 |
71
- | 0.0731 | 21.0 | 6846 | 4.8622 |
72
- | 0.0849 | 22.0 | 7172 | 4.8057 |
73
- | 0.1515 | 23.0 | 7498 | 4.9841 |
74
- | 0.2565 | 24.0 | 7824 | 5.3223 |
75
- | 0.2449 | 25.0 | 8150 | 4.3633 |
76
 
77
 
78
  ### Framework versions
79
 
80
  - PEFT 0.8.2
81
- - Transformers 4.37.2
82
  - Pytorch 2.0.1
83
  - Datasets 2.17.1
84
  - Tokenizers 0.15.2
 
16
 
17
  This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 4.3731
20
 
21
  ## Model description
22
 
 
41
  - seed: 42
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: constant
44
+ - num_epochs: 20
45
  - mixed_precision_training: Native AMP
46
 
47
  ### Training results
48
 
49
  | Training Loss | Epoch | Step | Validation Loss |
50
  |:-------------:|:-----:|:----:|:---------------:|
51
+ | 2.8286 | 1.0 | 326 | 2.7831 |
52
+ | 1.7135 | 2.0 | 652 | 2.9345 |
53
+ | 0.5324 | 3.0 | 978 | 3.2704 |
54
+ | 0.6332 | 4.0 | 1304 | 3.5215 |
55
+ | 0.4136 | 5.0 | 1630 | 3.6194 |
56
+ | 0.6789 | 6.0 | 1956 | 4.0601 |
57
+ | 0.3095 | 7.0 | 2282 | 3.9619 |
58
+ | 0.2139 | 8.0 | 2608 | 4.2931 |
59
+ | 0.2027 | 9.0 | 2934 | 4.3885 |
60
+ | 0.1184 | 10.0 | 3260 | 4.2185 |
61
+ | 0.1612 | 11.0 | 3586 | 4.2801 |
62
+ | 0.2609 | 12.0 | 3912 | 4.4705 |
63
+ | 0.1564 | 13.0 | 4238 | 4.7184 |
64
+ | 0.2344 | 14.0 | 4564 | 4.3517 |
65
+ | 0.4565 | 15.0 | 4890 | 4.7181 |
66
+ | 0.1623 | 16.0 | 5216 | 4.7855 |
67
+ | 0.2934 | 17.0 | 5542 | 5.5058 |
68
+ | 0.1151 | 18.0 | 5868 | 4.6761 |
69
+ | 0.178 | 19.0 | 6194 | 5.0001 |
70
+ | 0.1595 | 20.0 | 6520 | 4.3731 |
 
 
 
 
 
71
 
72
 
73
  ### Framework versions
74
 
75
  - PEFT 0.8.2
76
+ - Transformers 4.38.1
77
  - Pytorch 2.0.1
78
  - Datasets 2.17.1
79
  - Tokenizers 0.15.2
adapter_config.json CHANGED
@@ -18,17 +18,18 @@
18
  "embed_tokens"
19
  ],
20
  "peft_type": "LORA",
21
- "r": 64,
22
  "rank_pattern": {},
23
  "revision": null,
24
  "target_modules": [
25
- "down_proj",
26
  "gate_proj",
27
- "q_proj",
 
28
  "up_proj",
 
29
  "o_proj",
30
- "k_proj",
31
- "v_proj"
32
  ],
33
  "task_type": "CAUSAL_LM",
34
  "use_rslora": false
 
18
  "embed_tokens"
19
  ],
20
  "peft_type": "LORA",
21
+ "r": 16,
22
  "rank_pattern": {},
23
  "revision": null,
24
  "target_modules": [
 
25
  "gate_proj",
26
+ "down_proj",
27
+ "k_proj",
28
  "up_proj",
29
+ "v_proj",
30
  "o_proj",
31
+ "lm_head",
32
+ "q_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_rslora": false
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b6a2ae0adb81d32a5cecca68c3ff3ae18b9e1f27df002f0c66efe690f339ccb7
3
- size 3817075560
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7a90422abad5e331bbda477ca0b50e713d854c34e52466f6289c2cb1a094da5
3
+ size 1216475480
tokenizer_config.json CHANGED
@@ -1,6 +1,7 @@
1
  {
2
  "add_bos_token": true,
3
  "add_eos_token": false,
 
4
  "added_tokens_decoder": {
5
  "0": {
6
  "content": "<unk>",
 
1
  {
2
  "add_bos_token": true,
3
  "add_eos_token": false,
4
+ "add_prefix_space": true,
5
  "added_tokens_decoder": {
6
  "0": {
7
  "content": "<unk>",
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c666da4ebe994f7a208ee1a878a7481eaef585917b21ccec62afcd74678f4d85
3
- size 4283
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e565f3ecf98b05a32f4bfd167b09abd90ca6d3056b202ec6ef5130d64cf4405
3
+ size 4475