Chander-matrubhutam's workspace
Runs
5
Name
3 visualized
Updated
Oct 11 '24 22:40
Oct 11 '24 21:56
Oct 11 '24 21:37
Oct 11 '24 21:15
Oct 11 '24 20:53
State
Notes
User
Tags
Created
Runtime
Sweep
batch_size
eval_every_n_steps
learning_rate
max_eval_steps
max_steps
model_name
num_epochs
print_every_n_steps
seq_length
epoch
train/accuracy
train/loss
train/step_hz
train/step_time
train_step
Finished
wyler-zahm
Colab
14m
-
64
10
0.001
1
20
colab-llama-3.1-8B-Instruct-JAX
2
5
32
1
0.97119
0.17833
0.058303
17.15172
20
Finished
wyler-zahm
Colab
13m 42s
-
32
10
0.001
1
20
colab-llama-3.1-8B-Instruct-JAX
2
5
32
1
0.9707
0.16816
0.060292
16.58592
20
Finished
wyler-zahm
Colab
13m 45s
-
16
10
0.001
1
20
colab-llama-3.1-8B-Instruct-JAX
2
5
32
1
0.97461
0.14646
0.060433
16.54728
20
Finished
wyler-zahm
Colab
13m 32s
-
8
10
0.001
1
20
colab-llama-3.1-8B-Instruct-JAX
2
5
32
1
0.96875
0.19105
0.06108
16.37185
20
Finished
wyler-zahm
Colab
11m 37s
-
24
10
0.001
1
20
colab-llama-3.1-8B-Instruct-JAX
2
5
32
1
0.97266
0.1864
0.074071
13.50048
20
1-5
of 5