Hugging Face's logo

decryptellix
/
Llama-3.1-8B-LoRa-HP-check

PEFT
TensorBoard
Safetensors
Model card Files Metrics Community
Udith-Sandaruwan commited on
Commit
600dfec
·
verified ·
1 Parent(s): e5919ec

Organize logs inside session_logs/

Browse files
Files changed (5) hide show
  1. session_logs/evaluation_results.json +72 -0
  2. session_logs/hyperparameter_results.csv +3 -0
  3. session_logs/logs/events.out.tfevents.1741581719.1cb3cfe4f1cc.3544.0 +3 -0
  4. session_logs/logs/events.out.tfevents.1741585711.1cb3cfe4f1cc.3544.1 +3 -0
  5. session_logs/lora_finetuning_hp.log +11 -0
session_logs/evaluation_results.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "1": {
3
+ "lora_alpha": 23,
4
+ "lora_dropout": 0.25385249450462455,
5
+ "learning_rate": 0.0002521627789110728,
6
+ "gradient_accumulation_steps": 3,
7
+ "lora_rank": 13,
8
+ "eval_results": {
9
+ "meteor_scores": {
10
+ "meteor": 0.19398200858217152
11
+ },
12
+ "rouge_scores": {
13
+ "rouge1": 0.04557065217391304,
14
+ "rouge2": 0.008522727272727272,
15
+ "rougeL": 0.0453968253968254,
16
+ "rougeLsum": 0.045932539682539686
17
+ },
18
+ "bleu_scores": {
19
+ "bleu": 0.05877275210755556,
20
+ "precisions": [
21
+ 0.10538218000904569,
22
+ 0.06435417617526244,
23
+ 0.04790419161676647,
24
+ 0.03672710367271037
25
+ ],
26
+ "brevity_penalty": 1.0,
27
+ "length_ratio": 2.660649819494585,
28
+ "translation_length": 2211,
29
+ "reference_length": 831
30
+ },
31
+ "perplexity": 5552.1064453125
32
+ }
33
+ },
34
+ "2": {
35
+ "lora_alpha": 18,
36
+ "lora_dropout": 0.11479151282864085,
37
+ "learning_rate": 0.0005195992689054757,
38
+ "gradient_accumulation_steps": 3,
39
+ "lora_rank": 22,
40
+ "eval_results": {
41
+ "meteor_scores": {
42
+ "meteor": 0.2074470076536067
43
+ },
44
+ "rouge_scores": {
45
+ "rouge1": 0.10333333333333332,
46
+ "rouge2": 0.05833333333333333,
47
+ "rougeL": 0.1,
48
+ "rougeLsum": 0.1
49
+ },
50
+ "bleu_scores": {
51
+ "bleu": 0.06657896994755348,
52
+ "precisions": [
53
+ 0.12014314928425358,
54
+ 0.07231404958677685,
55
+ 0.054279749478079335,
56
+ 0.041666666666666664
57
+ ],
58
+ "brevity_penalty": 1.0,
59
+ "length_ratio": 2.3537906137184117,
60
+ "translation_length": 1956,
61
+ "reference_length": 831
62
+ },
63
+ "perplexity": 6999.1513671875
64
+ }
65
+ },
66
+ "best_param": {
67
+ "lora_alpha": 23,
68
+ "lora_dropout": 0.25385249450462455,
69
+ "learning_rate": 0.0002521627789110728,
70
+ "gradient_accumulation_steps": 3
71
+ }
72
+ }
session_logs/hyperparameter_results.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ lora_alpha,lora_dropout,learning_rate,lora_rank,gradient_accumulation_steps,eval_results
2
+ 23,0.25385249450462455,0.0002521627789110728,13,3,"{'meteor_scores': {'meteor': 0.19398200858217152}, 'rouge_scores': {'rouge1': 0.04557065217391304, 'rouge2': 0.008522727272727272, 'rougeL': 0.0453968253968254, 'rougeLsum': 0.045932539682539686}, 'bleu_scores': {'bleu': 0.05877275210755556, 'precisions': [0.10538218000904569, 0.06435417617526244, 0.04790419161676647, 0.03672710367271037], 'brevity_penalty': 1.0, 'length_ratio': 2.660649819494585, 'translation_length': 2211, 'reference_length': 831}, 'perplexity': 5552.1064453125}"
3
+ 18,0.11479151282864085,0.0005195992689054757,22,3,"{'meteor_scores': {'meteor': 0.2074470076536067}, 'rouge_scores': {'rouge1': 0.10333333333333332, 'rouge2': 0.05833333333333333, 'rougeL': 0.1, 'rougeLsum': 0.1}, 'bleu_scores': {'bleu': 0.06657896994755348, 'precisions': [0.12014314928425358, 0.07231404958677685, 0.054279749478079335, 0.041666666666666664], 'brevity_penalty': 1.0, 'length_ratio': 2.3537906137184117, 'translation_length': 1956, 'reference_length': 831}, 'perplexity': 6999.1513671875}"
session_logs/logs/events.out.tfevents.1741581719.1cb3cfe4f1cc.3544.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13f6532beee9c6d26d443842147e178cd95225f2e36198fae9c4385621cfdfb7
3
+ size 7412
session_logs/logs/events.out.tfevents.1741585711.1cb3cfe4f1cc.3544.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a96fe2fc7bc2e652c9772bd56d0b2732ce4d89c27b527e44be31bbfc8d340488
3
+ size 7412
session_logs/lora_finetuning_hp.log ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2025-03-10 04:28:23,982 - Logging initialized for session: e5af214f-503b-4ff4-a6d4-d7bf43ab68d8
2
+ 2025-03-10 04:29:08,667 - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
3
+ 2025-03-10 04:30:58,611 - Training with params: lora_alpha=23, lora_dropout=0.25385249450462455, learning_rate=0.0002521627789110728, gradient_accumulation_steps=3
4
+ 2025-03-10 05:36:55,205 - Using default tokenizer.
5
+ 2025-03-10 05:37:00,453 - Evaluation Results: {'meteor_scores': {'meteor': 0.19398200858217152}, 'rouge_scores': {'rouge1': 0.04557065217391304, 'rouge2': 0.008522727272727272, 'rougeL': 0.0453968253968254, 'rougeLsum': 0.045932539682539686}, 'bleu_scores': {'bleu': 0.05877275210755556, 'precisions': [0.10538218000904569, 0.06435417617526244, 0.04790419161676647, 0.03672710367271037], 'brevity_penalty': 1.0, 'length_ratio': 2.660649819494585, 'translation_length': 2211, 'reference_length': 831}, 'perplexity': 5552.1064453125}
6
+ 2025-03-10 05:37:00,678 - Training with params: lora_alpha=18, lora_dropout=0.11479151282864085, learning_rate=0.0005195992689054757, gradient_accumulation_steps=3
7
+ 2025-03-10 05:37:00,693 - Already found a `peft_config` attribute in the model. This will lead to having multiple adapters in the model. Make sure to know what you are doing!
8
+ 2025-03-10 06:41:50,465 - Using default tokenizer.
9
+ 2025-03-10 06:41:53,425 - Evaluation Results: {'meteor_scores': {'meteor': 0.2074470076536067}, 'rouge_scores': {'rouge1': 0.10333333333333332, 'rouge2': 0.05833333333333333, 'rougeL': 0.1, 'rougeLsum': 0.1}, 'bleu_scores': {'bleu': 0.06657896994755348, 'precisions': [0.12014314928425358, 0.07231404958677685, 0.054279749478079335, 0.041666666666666664], 'brevity_penalty': 1.0, 'length_ratio': 2.3537906137184117, 'translation_length': 1956, 'reference_length': 831}, 'perplexity': 6999.1513671875}
10
+ 2025-03-10 06:41:53,431 - Best hyperparameters: {'gradient_accumulation_steps': 3.224340351235237, 'learning_rate': 0.0002521627789110728, 'lora_alpha': 22.976944309938723, 'lora_dropout': 0.25385249450462455, 'lora_rank': 12.725204870446019}
11
+ 2025-03-10 06:41:53,493 - Already found a `peft_config` attribute in the model. This will lead to having multiple adapters in the model. Make sure to know what you are doing!