End of training
Browse files- README.md +17 -1
- all_results.json +13 -0
- config.json +1 -1
- eval_results.json +16 -0
README.md
CHANGED
@@ -2,9 +2,15 @@
|
|
2 |
license: apache-2.0
|
3 |
base_model: Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1
|
4 |
tags:
|
|
|
5 |
- trl
|
6 |
- dpo
|
7 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
8 |
model-index:
|
9 |
- name: biomistral-7b-wo-kqa_golden-iter-dpo-step2
|
10 |
results: []
|
@@ -15,7 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# biomistral-7b-wo-kqa_golden-iter-dpo-step2
|
17 |
|
18 |
-
This model is a fine-tuned version of [Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1](https://huggingface.co/Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1) on
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
## Model description
|
21 |
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1
|
4 |
tags:
|
5 |
+
- alignment-handbook
|
6 |
- trl
|
7 |
- dpo
|
8 |
- generated_from_trainer
|
9 |
+
- trl
|
10 |
+
- dpo
|
11 |
+
- generated_from_trainer
|
12 |
+
datasets:
|
13 |
+
- HuggingFaceH4/ultrafeedback_binarized
|
14 |
model-index:
|
15 |
- name: biomistral-7b-wo-kqa_golden-iter-dpo-step2
|
16 |
results: []
|
|
|
21 |
|
22 |
# biomistral-7b-wo-kqa_golden-iter-dpo-step2
|
23 |
|
24 |
+
This model is a fine-tuned version of [Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1](https://huggingface.co/Minbyul/biomistral-7b-wo-kqa_golden-iter-dpo-step1) on the HuggingFaceH4/ultrafeedback_binarized dataset.
|
25 |
+
It achieves the following results on the evaluation set:
|
26 |
+
- Loss: 0.6909
|
27 |
+
- Rewards/chosen: 0.0063
|
28 |
+
- Rewards/rejected: 0.0057
|
29 |
+
- Rewards/accuracies: 0.5625
|
30 |
+
- Rewards/margins: 0.0006
|
31 |
+
- Logps/rejected: -193.8717
|
32 |
+
- Logps/chosen: -168.4928
|
33 |
+
- Logits/rejected: -2.2060
|
34 |
+
- Logits/chosen: -2.9391
|
35 |
|
36 |
## Model description
|
37 |
|
all_results.json
CHANGED
@@ -1,5 +1,18 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
"train_loss": 0.680875283606509,
|
4 |
"train_runtime": 1044.3353,
|
5 |
"train_samples": 6013,
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_logits/chosen": -2.9391283988952637,
|
4 |
+
"eval_logits/rejected": -2.2059857845306396,
|
5 |
+
"eval_logps/chosen": -168.49281311035156,
|
6 |
+
"eval_logps/rejected": -193.87171936035156,
|
7 |
+
"eval_loss": 0.6908513903617859,
|
8 |
+
"eval_rewards/accuracies": 0.5625,
|
9 |
+
"eval_rewards/chosen": 0.00629099877551198,
|
10 |
+
"eval_rewards/margins": 0.0006325023132376373,
|
11 |
+
"eval_rewards/rejected": 0.0056584966368973255,
|
12 |
+
"eval_runtime": 20.5327,
|
13 |
+
"eval_samples": 231,
|
14 |
+
"eval_samples_per_second": 11.25,
|
15 |
+
"eval_steps_per_second": 0.39,
|
16 |
"train_loss": 0.680875283606509,
|
17 |
"train_runtime": 1044.3353,
|
18 |
"train_samples": 6013,
|
config.json
CHANGED
@@ -21,6 +21,6 @@
|
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
"transformers_version": "4.38.2",
|
24 |
-
"use_cache":
|
25 |
"vocab_size": 32000
|
26 |
}
|
|
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
"transformers_version": "4.38.2",
|
24 |
+
"use_cache": true,
|
25 |
"vocab_size": 32000
|
26 |
}
|
eval_results.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_logits/chosen": -2.9391283988952637,
|
4 |
+
"eval_logits/rejected": -2.2059857845306396,
|
5 |
+
"eval_logps/chosen": -168.49281311035156,
|
6 |
+
"eval_logps/rejected": -193.87171936035156,
|
7 |
+
"eval_loss": 0.6908513903617859,
|
8 |
+
"eval_rewards/accuracies": 0.5625,
|
9 |
+
"eval_rewards/chosen": 0.00629099877551198,
|
10 |
+
"eval_rewards/margins": 0.0006325023132376373,
|
11 |
+
"eval_rewards/rejected": 0.0056584966368973255,
|
12 |
+
"eval_runtime": 20.5327,
|
13 |
+
"eval_samples": 231,
|
14 |
+
"eval_samples_per_second": 11.25,
|
15 |
+
"eval_steps_per_second": 0.39
|
16 |
+
}
|