Hugging Face's logo

Shahradmz
/
OLMo-1B-hf-DPO-constitution-full-2

Transformers
Safetensors
Generated from Trainer
trl
dpo
Inference Endpoints
Model card Files Community
Shahradmz commited on
Commit
2203d4e
•
1 Parent(s): 50e8f44

Model save

Browse files
Files changed (2) hide show
  1. all_results.json +15 -0
  2. eval_results.json +15 -0
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9999343530493009,
3
+ "eval_logits/chosen": -0.8756721019744873,
4
+ "eval_logits/rejected": -0.9693624377250671,
5
+ "eval_logps/chosen": -103.03213500976562,
6
+ "eval_logps/rejected": -254.58119201660156,
7
+ "eval_loss": 0.0005562438163906336,
8
+ "eval_rewards/accuracies": 1.0,
9
+ "eval_rewards/chosen": -3.5039849281311035,
10
+ "eval_rewards/margins": 14.41648006439209,
11
+ "eval_rewards/rejected": -17.92046546936035,
12
+ "eval_runtime": 161.905,
13
+ "eval_samples_per_second": 33.211,
14
+ "eval_steps_per_second": 4.157
15
+ }
eval_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9999343530493009,
3
+ "eval_logits/chosen": -0.8756721019744873,
4
+ "eval_logits/rejected": -0.9693624377250671,
5
+ "eval_logps/chosen": -103.03213500976562,
6
+ "eval_logps/rejected": -254.58119201660156,
7
+ "eval_loss": 0.0005562438163906336,
8
+ "eval_rewards/accuracies": 1.0,
9
+ "eval_rewards/chosen": -3.5039849281311035,
10
+ "eval_rewards/margins": 14.41648006439209,
11
+ "eval_rewards/rejected": -17.92046546936035,
12
+ "eval_runtime": 161.905,
13
+ "eval_samples_per_second": 33.211,
14
+ "eval_steps_per_second": 4.157
15
+ }