End of training
Browse files- README.md +14 -8
- 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: alignment-handbook/zephyr-7b-sft-full
|
4 |
tags:
|
|
|
5 |
- trl
|
6 |
- dpo
|
7 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
8 |
model-index:
|
9 |
- name: zephyr-7b-dpo-full-ultrabin-reward-scale-05
|
10 |
results: []
|
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# zephyr-7b-dpo-full-ultrabin-reward-scale-05
|
17 |
|
18 |
-
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on
|
19 |
It achieves the following results on the evaluation set:
|
20 |
- Loss: 0.5419
|
21 |
-
- Rewards/chosen: -2.
|
22 |
-
- Rewards/rejected: -3.
|
23 |
- Rewards/accuracies: 0.7812
|
24 |
-
- Rewards/margins: 1.
|
25 |
-
- Logps/rejected: -617.
|
26 |
-
- Logps/chosen: -469.
|
27 |
-
- Logits/rejected: 3.
|
28 |
-
- Logits/chosen: 2.
|
29 |
|
30 |
## Model description
|
31 |
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: alignment-handbook/zephyr-7b-sft-full
|
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: zephyr-7b-dpo-full-ultrabin-reward-scale-05
|
16 |
results: []
|
|
|
21 |
|
22 |
# zephyr-7b-dpo-full-ultrabin-reward-scale-05
|
23 |
|
24 |
+
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the HuggingFaceH4/ultrafeedback_binarized dataset.
|
25 |
It achieves the following results on the evaluation set:
|
26 |
- Loss: 0.5419
|
27 |
+
- Rewards/chosen: -2.0652
|
28 |
+
- Rewards/rejected: -3.5533
|
29 |
- Rewards/accuracies: 0.7812
|
30 |
+
- Rewards/margins: 1.4880
|
31 |
+
- Logps/rejected: -617.9895
|
32 |
+
- Logps/chosen: -469.1545
|
33 |
+
- Logits/rejected: 3.0313
|
34 |
+
- Logits/chosen: 2.1931
|
35 |
|
36 |
## Model description
|
37 |
|
all_results.json
CHANGED
@@ -1,5 +1,18 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
"total_flos": 0.0,
|
4 |
"train_loss": 0.5571172007955767,
|
5 |
"train_runtime": 12724.9946,
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_logits/chosen": 2.193073272705078,
|
4 |
+
"eval_logits/rejected": 3.0313031673431396,
|
5 |
+
"eval_logps/chosen": -469.1545104980469,
|
6 |
+
"eval_logps/rejected": -617.989501953125,
|
7 |
+
"eval_loss": 0.5419281721115112,
|
8 |
+
"eval_rewards/accuracies": 0.78125,
|
9 |
+
"eval_rewards/chosen": -2.0652458667755127,
|
10 |
+
"eval_rewards/margins": 1.4880279302597046,
|
11 |
+
"eval_rewards/rejected": -3.5532736778259277,
|
12 |
+
"eval_runtime": 103.7243,
|
13 |
+
"eval_samples": 2000,
|
14 |
+
"eval_samples_per_second": 19.282,
|
15 |
+
"eval_steps_per_second": 0.309,
|
16 |
"total_flos": 0.0,
|
17 |
"train_loss": 0.5571172007955767,
|
18 |
"train_runtime": 12724.9946,
|
config.json
CHANGED
@@ -22,6 +22,6 @@
|
|
22 |
"tie_word_embeddings": false,
|
23 |
"torch_dtype": "bfloat16",
|
24 |
"transformers_version": "4.44.0.dev0",
|
25 |
-
"use_cache":
|
26 |
"vocab_size": 32000
|
27 |
}
|
|
|
22 |
"tie_word_embeddings": false,
|
23 |
"torch_dtype": "bfloat16",
|
24 |
"transformers_version": "4.44.0.dev0",
|
25 |
+
"use_cache": true,
|
26 |
"vocab_size": 32000
|
27 |
}
|
eval_results.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_logits/chosen": 2.193073272705078,
|
4 |
+
"eval_logits/rejected": 3.0313031673431396,
|
5 |
+
"eval_logps/chosen": -469.1545104980469,
|
6 |
+
"eval_logps/rejected": -617.989501953125,
|
7 |
+
"eval_loss": 0.5419281721115112,
|
8 |
+
"eval_rewards/accuracies": 0.78125,
|
9 |
+
"eval_rewards/chosen": -2.0652458667755127,
|
10 |
+
"eval_rewards/margins": 1.4880279302597046,
|
11 |
+
"eval_rewards/rejected": -3.5532736778259277,
|
12 |
+
"eval_runtime": 103.7243,
|
13 |
+
"eval_samples": 2000,
|
14 |
+
"eval_samples_per_second": 19.282,
|
15 |
+
"eval_steps_per_second": 0.309
|
16 |
+
}
|