Hugging Face's logo

MaRyAm1295
/
Llama-3.2-3B-KAM

PEFT
Safetensors
llama
trl
sft
Llama
Generated from Trainer
4-bit precision
bitsandbytes
Model card Files Community
MaRyAm1295 commited on
Commit
3fa184d
·
verified ·
1 Parent(s): ebcbb99

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +25 -25
README.md CHANGED
@@ -49,31 +49,31 @@ The following hyperparameters were used during training:
49
  - mixed_precision_training: Native AMP
50
 
51
  ### Training results
52
- ### Step    Training Loss
53
- - 50    2.436700
54
- - 100    2.103400
55
- - 150    2.048900
56
- - 200    2.041700
57
- - 250    2.002900
58
- - 300    1.991700
59
- - 350    1.977400
60
- - 400    1.974500
61
- - 450    1.945000
62
- - 500    1.951100
63
- - 550    1.950700
64
- - 600    1.943000
65
- - 650    1.927900
66
- - 700    1.920900
67
- - 750    1.903400
68
- - 800    1.896000
69
- - 850    1.910800
70
- - 900    1.904600
71
- - 950    1.918100
72
- - 1000    1.911500
73
- - 1050    1.909100
74
- - 1100    1.928900
75
- - 1150    1.896100
76
- - 1200    1.876700
77
 
78
 
79
 
 
49
  - mixed_precision_training: Native AMP
50
 
51
  ### Training results
52
+ #### Step    Training Loss
53
+ - 50     2.436700
54
+ - 100     2.103400
55
+ - 150     2.048900
56
+ - 200     2.041700
57
+ - 250     2.002900
58
+ - 300     1.991700
59
+ - 350     1.977400
60
+ - 400     1.974500
61
+ - 450     1.945000
62
+ - 500     1.951100
63
+ - 550     1.950700
64
+ - 600     1.943000
65
+ - 650     1.927900
66
+ - 700     1.920900
67
+ - 750     1.903400
68
+ - 800     1.896000
69
+ - 850     1.910800
70
+ - 900     1.904600
71
+ - 950     1.918100
72
+ - 1000     1.911500
73
+ - 1050     1.909100
74
+ - 1100     1.928900
75
+ - 1150     1.896100
76
+ - 1200     1.876700
77
 
78
 
79