Hugging Face's logo

thrunlab
/
mistral_sparse_80_percent_boolq_1000

Transformers
Safetensors
mistral
trl
sft
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
lukeleeai commited on
Commit
9232c9f
•
1 Parent(s): aac50a5

Training in progress, step 50

Browse files
Files changed (7) hide show
  1. README.md +32 -24
  2. adapter_config.json +2 -2
  3. adapter_model.safetensors +1 -1
  4. model-00001-of-00003.safetensors +1 -1
  5. model-00002-of-00003.safetensors +1 -1
  6. model-00003-of-00003.safetensors +1 -1
  7. training_args.bin +1 -1
README.md CHANGED
@@ -19,8 +19,8 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  This model is a fine-tuned version of [](https://huggingface.co/) on the super_glue dataset.
21
  It achieves the following results on the evaluation set:
22
- - Loss: 0.3573
23
- - Accuracy: 0.8580
24
 
25
  ## Model description
26
 
@@ -42,7 +42,7 @@ The following hyperparameters were used during training:
42
  - learning_rate: 1e-05
43
  - train_batch_size: 2
44
  - eval_batch_size: 4
45
- - seed: 0
46
  - distributed_type: multi-GPU
47
  - num_devices: 2
48
  - gradient_accumulation_steps: 2
@@ -56,27 +56,35 @@ The following hyperparameters were used during training:
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
58
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
59
- | 0.6397 | 0.05 | 50 | 0.4990 | 0.7802 |
60
- | 0.4151 | 0.1 | 100 | 0.4801 | 0.7972 |
61
- | 0.1805 | 0.15 | 150 | 0.4870 | 0.7965 |
62
- | 0.4189 | 0.2 | 200 | 0.4392 | 0.8240 |
63
- | 0.2601 | 0.25 | 250 | 0.4282 | 0.8311 |
64
- | 0.3197 | 0.3 | 300 | 0.4401 | 0.8375 |
65
- | 0.1839 | 0.35 | 350 | 0.4113 | 0.8382 |
66
- | 0.4575 | 0.4 | 400 | 0.3906 | 0.8417 |
67
- | 0.5272 | 0.45 | 450 | 0.3932 | 0.8466 |
68
- | 0.2914 | 0.5 | 500 | 0.4045 | 0.8445 |
69
- | 0.4161 | 0.55 | 550 | 0.4602 | 0.8382 |
70
- | 0.245 | 0.6 | 600 | 0.3853 | 0.8495 |
71
- | 0.4237 | 0.65 | 650 | 0.3665 | 0.8537 |
72
- | 0.1947 | 0.7 | 700 | 0.3632 | 0.8601 |
73
- | 0.4434 | 0.75 | 750 | 0.3511 | 0.8601 |
74
- | 0.4321 | 0.8 | 800 | 0.3905 | 0.8615 |
75
- | 0.2262 | 0.85 | 850 | 0.3542 | 0.8587 |
76
- | 0.41 | 0.9 | 900 | 0.3773 | 0.8572 |
77
- | 0.585 | 0.95 | 950 | 0.3701 | 0.8572 |
78
- | 0.3416 | 1.0 | 1000 | 0.3616 | 0.8608 |
79
- | 0.9198 | 1.05 | 1050 | 0.3589 | 0.8580 |
 
 
 
 
 
 
 
 
80
 
81
 
82
  ### Framework versions
 
19
 
20
  This model is a fine-tuned version of [](https://huggingface.co/) on the super_glue dataset.
21
  It achieves the following results on the evaluation set:
22
+ - Loss: 0.3412
23
+ - Accuracy: 0.8678
24
 
25
  ## Model description
26
 
 
42
  - learning_rate: 1e-05
43
  - train_batch_size: 2
44
  - eval_batch_size: 4
45
+ - seed: 1
46
  - distributed_type: multi-GPU
47
  - num_devices: 2
48
  - gradient_accumulation_steps: 2
 
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
58
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
59
+ | 0.4585 | 0.05 | 50 | 0.4872 | 0.7760 |
60
+ | 0.3385 | 0.1 | 100 | 0.4658 | 0.7880 |
61
+ | 0.8423 | 0.15 | 150 | 0.4830 | 0.8113 |
62
+ | 0.4092 | 0.2 | 200 | 0.4115 | 0.8269 |
63
+ | 0.4483 | 0.25 | 250 | 0.4318 | 0.8120 |
64
+ | 0.3319 | 0.3 | 300 | 0.4008 | 0.8346 |
65
+ | 0.3072 | 0.35 | 350 | 0.4538 | 0.8346 |
66
+ | 0.2787 | 0.4 | 400 | 0.4432 | 0.8332 |
67
+ | 0.2965 | 0.45 | 450 | 0.3921 | 0.8403 |
68
+ | 0.2684 | 0.5 | 500 | 0.3777 | 0.8445 |
69
+ | 0.2543 | 0.55 | 550 | 0.3793 | 0.8481 |
70
+ | 0.196 | 0.6 | 600 | 0.3703 | 0.8509 |
71
+ | 0.2828 | 0.65 | 650 | 0.3910 | 0.8509 |
72
+ | 0.2073 | 0.7 | 700 | 0.3813 | 0.8587 |
73
+ | 0.2733 | 0.75 | 750 | 0.3841 | 0.8650 |
74
+ | 0.3351 | 0.8 | 800 | 0.3658 | 0.8643 |
75
+ | 0.126 | 0.85 | 850 | 0.3713 | 0.8643 |
76
+ | 0.3351 | 0.9 | 900 | 0.3457 | 0.8594 |
77
+ | 0.3614 | 0.95 | 950 | 0.3645 | 0.8629 |
78
+ | 0.3383 | 1.0 | 1000 | 0.3809 | 0.8615 |
79
+ | 0.2019 | 1.05 | 1050 | 0.4630 | 0.8707 |
80
+ | 0.5042 | 1.1 | 1100 | 0.3724 | 0.8657 |
81
+ | 0.2159 | 1.15 | 1150 | 0.3449 | 0.8643 |
82
+ | 0.4469 | 1.2 | 1200 | 0.3555 | 0.8693 |
83
+ | 0.6519 | 1.25 | 1250 | 0.4500 | 0.8686 |
84
+ | 0.0631 | 1.3 | 1300 | 0.4127 | 0.8678 |
85
+ | 0.2844 | 1.35 | 1350 | 0.3950 | 0.8721 |
86
+ | 0.3926 | 1.4 | 1400 | 0.3742 | 0.8714 |
87
+ | 0.0955 | 1.45 | 1450 | 0.3810 | 0.8707 |
88
 
89
 
90
  ### Framework versions
adapter_config.json CHANGED
@@ -19,10 +19,10 @@
19
  "rank_pattern": {},
20
  "revision": null,
21
  "target_modules": [
22
- "v_proj",
23
  "q_proj",
24
  "gate_proj",
25
- "down_proj"
 
26
  ],
27
  "task_type": "CAUSAL_LM"
28
  }
 
19
  "rank_pattern": {},
20
  "revision": null,
21
  "target_modules": [
 
22
  "q_proj",
23
  "gate_proj",
24
+ "down_proj",
25
+ "v_proj"
26
  ],
27
  "task_type": "CAUSAL_LM"
28
  }
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1449e751b1ac8d88a59a626967d40884b6ca2f12a5f83ade084bbbe51093d66f
3
  size 205555592
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46e79effb4eb07ba65dc4dd45dd352461632d89ecb1d76915ce809c8f5ddaff3
3
  size 205555592
model-00001-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6bff80c30292073d5da33263b8373c95248e2e6bceec40f4c94a791df5e2a6ca
3
  size 4943163992
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cb4f6ee8818d7f3a150dd0ed9657780fddf17bc442f642063aa6515bfefefc0
3
  size 4943163992
model-00002-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4c28d3e73304f62c2444300b2d5c4c6935fdf0de0375d13c23370038735eac7c
3
  size 4999821144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03cd1ce2b0f84e5635ce6b03704c75b34f8cc2a11eb1e546f95428fef02937f0
3
  size 4999821144
model-00003-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6334877f77cea916d690b02a81cc8e1055e711ab307818eacfb11dc9a6298627
3
  size 4540517840
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e3ab5f6637f696df3f15f4ba11068bfebad8eac84f0e7a4d2bde70a795aa5c0e
3
  size 4540517840
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2f3805618f1e860e17074772a001b83dfaa4c6ebfa170013656e3713f1e01a0b
3
  size 6008
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b16e4e572a5397d75c40edc507c121be0234f09479c08970ce847a154b86ed82
3
  size 6008