Commit
•
e4d0aae
1
Parent(s):
577946f
End of training
Browse files- README.md +16 -10
- all_results.json +13 -0
- config.json +1 -1
- eval_results.json +16 -0
- runs/Jun07_09-08-56_poseidon/events.out.tfevents.1717836367.poseidon.3198658.1 +3 -0
README.md
CHANGED
@@ -2,9 +2,15 @@
|
|
2 |
license: apache-2.0
|
3 |
base_model: martimfasantos/tinyllama-1.1b-sum-sft-full
|
4 |
tags:
|
|
|
5 |
- trl
|
6 |
- dpo
|
7 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
8 |
model-index:
|
9 |
- name: tinyllama-1.1b-sum-dpo-full_LR5e-7_3epochs
|
10 |
results: []
|
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# tinyllama-1.1b-sum-dpo-full_LR5e-7_3epochs
|
17 |
|
18 |
-
This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full) on
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss: 0.
|
21 |
-
- Rewards/chosen: -2.
|
22 |
-
- Rewards/rejected: -3.
|
23 |
-
- Rewards/accuracies: 0.
|
24 |
-
- Rewards/margins: 0.
|
25 |
-
- Logps/rejected: -404.
|
26 |
-
- Logps/chosen: -
|
27 |
-
- Logits/rejected: -1.
|
28 |
-
- Logits/chosen: -2.
|
29 |
|
30 |
## Model description
|
31 |
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: martimfasantos/tinyllama-1.1b-sum-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 |
+
- openai/summarize_from_feedback
|
14 |
model-index:
|
15 |
- name: tinyllama-1.1b-sum-dpo-full_LR5e-7_3epochs
|
16 |
results: []
|
|
|
21 |
|
22 |
# tinyllama-1.1b-sum-dpo-full_LR5e-7_3epochs
|
23 |
|
24 |
+
This model is a fine-tuned version of [martimfasantos/tinyllama-1.1b-sum-sft-full](https://huggingface.co/martimfasantos/tinyllama-1.1b-sum-sft-full) on the openai/summarize_from_feedback dataset.
|
25 |
It achieves the following results on the evaluation set:
|
26 |
+
- Loss: 0.7099
|
27 |
+
- Rewards/chosen: -2.8601
|
28 |
+
- Rewards/rejected: -3.4154
|
29 |
+
- Rewards/accuracies: 0.6320
|
30 |
+
- Rewards/margins: 0.5553
|
31 |
+
- Logps/rejected: -404.2897
|
32 |
+
- Logps/chosen: -345.0273
|
33 |
+
- Logits/rejected: -1.9822
|
34 |
+
- Logits/chosen: -2.0068
|
35 |
|
36 |
## Model description
|
37 |
|
all_results.json
CHANGED
@@ -1,5 +1,18 @@
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
"total_flos": 0.0,
|
4 |
"train_loss": 0.49131362667692796,
|
5 |
"train_runtime": 84120.8122,
|
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
+
"eval_logits/chosen": -2.0068085193634033,
|
4 |
+
"eval_logits/rejected": -1.9822146892547607,
|
5 |
+
"eval_logps/chosen": -345.02728271484375,
|
6 |
+
"eval_logps/rejected": -404.28973388671875,
|
7 |
+
"eval_loss": 0.7099329233169556,
|
8 |
+
"eval_rewards/accuracies": 0.6319702863693237,
|
9 |
+
"eval_rewards/chosen": -2.8601176738739014,
|
10 |
+
"eval_rewards/margins": 0.5552833080291748,
|
11 |
+
"eval_rewards/rejected": -3.4154014587402344,
|
12 |
+
"eval_runtime": 360.569,
|
13 |
+
"eval_samples": 4304,
|
14 |
+
"eval_samples_per_second": 11.937,
|
15 |
+
"eval_steps_per_second": 1.492,
|
16 |
"total_flos": 0.0,
|
17 |
"train_loss": 0.49131362667692796,
|
18 |
"train_runtime": 84120.8122,
|
config.json
CHANGED
@@ -24,6 +24,6 @@
|
|
24 |
"tie_word_embeddings": false,
|
25 |
"torch_dtype": "float32",
|
26 |
"transformers_version": "4.41.2",
|
27 |
-
"use_cache":
|
28 |
"vocab_size": 32000
|
29 |
}
|
|
|
24 |
"tie_word_embeddings": false,
|
25 |
"torch_dtype": "float32",
|
26 |
"transformers_version": "4.41.2",
|
27 |
+
"use_cache": true,
|
28 |
"vocab_size": 32000
|
29 |
}
|
eval_results.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 3.0,
|
3 |
+
"eval_logits/chosen": -2.0068085193634033,
|
4 |
+
"eval_logits/rejected": -1.9822146892547607,
|
5 |
+
"eval_logps/chosen": -345.02728271484375,
|
6 |
+
"eval_logps/rejected": -404.28973388671875,
|
7 |
+
"eval_loss": 0.7099329233169556,
|
8 |
+
"eval_rewards/accuracies": 0.6319702863693237,
|
9 |
+
"eval_rewards/chosen": -2.8601176738739014,
|
10 |
+
"eval_rewards/margins": 0.5552833080291748,
|
11 |
+
"eval_rewards/rejected": -3.4154014587402344,
|
12 |
+
"eval_runtime": 360.569,
|
13 |
+
"eval_samples": 4304,
|
14 |
+
"eval_samples_per_second": 11.937,
|
15 |
+
"eval_steps_per_second": 1.492
|
16 |
+
}
|
runs/Jun07_09-08-56_poseidon/events.out.tfevents.1717836367.poseidon.3198658.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6b1d1167a242fd59d22039df9dcca3d4ec15190211b5658f5938e55a524c52c4
|
3 |
+
size 841
|