Capecape's workspace
Runs
398
Name
398 visualized
Tags
batch_size
gradient_accumulation_steps
lr
effective_batch_size
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
64
0.00002
32
7b
hf_sft
1
32
0.00002
32
7b
hf_sft
1
32
0.00002
32
7b
hf_sft
2
32
0.00002
32
7b
hf_sft
2
32
0.00002
32
7b
hf_sft
2
32
0.00002
32
7b
hf_sft
1
32
0.00002
32
7b
hf_sft
1
32
0.00002
32
7b
hf_sft
1
32
0.00002
32
7b
hf_sft
1
32
0.00002
32
7b
hf_sft
2
32
0.00002
32
7b
hf_sft
2
32
0.00002
32
7b
hf_sft
2
32
0.00002
32
State
Notes
User
Created
Runtime
Sweep
dataset_name
epoch_sz
epochs
freeze_embed
gradient_checkpointing
log_every
model_id
mom
n_cut
n_freeze
precision
save_model
total_params
trainable_params
_name_or_path
adafactor
adam_beta1
adam_beta2
adam_epsilon
add_cross_attention
architectures
attention_bias
auto_find_batch_size
bf16
bf16_full_eval
bos_token_id
chunk_size_feed_forward
dataloader_drop_last
dataloader_num_workers
dataloader_pin_memory
ddp_timeout
debug
disable_tqdm
diversity_penalty
do_eval
do_predict
do_sample
do_train
early_stopping
encoder_no_repeat_ngram_size
eos_token_id
eval_delay
eval_every
eval_steps
evaluation_strategy
Finished
-
capecape
11m 19s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
31m 48s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
16m 47s
-
-
-
true
False
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Failed
-
capecape
32s
-
-
-
true
False
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
10m 3s
-
-
-
true
False
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
56m 54s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
41m 4s
-
-
-
true
False
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Failed
-
capecape
33s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Failed
-
capecape
3h 8s
-
-
-
-
true
false
-
meta-llama/Llama-2-7b-hf
-
-
all
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
11m 44s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
21m 54s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
10m 41s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
32m 37s
-
-
-
-
true
false
-
meta-llama/Llama-2-7b-hf
-
-
all
-
true
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Killed
-
capecape
1m 12s
-
-
-
-
true
false
-
meta-llama/Llama-2-7b-hf
-
-
all
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
24m 18s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
3m 5s
-
-
-
-
true
false
-
meta-llama/Llama-2-7b-hf
-
-
all
-
true
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
4m 9s
-
-
-
-
true
false
-
meta-llama/Llama-2-7b-hf
-
-
all
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Finished
-
capecape
53m 47s
-
-
-
true
True
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Killed
-
capecape
1m 7s
-
-
-
-
true
false
-
meta-llama/Llama-2-7b-hf
-
-
all
-
true
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
Failed
-
capecape
18s
-
-
-
true
False
-
meta-llama/Llama-2-7b-hf
-
-
24
-
-
-
-
meta-llama/Llama-2-7b-hf
false
0.9
0.999
1.0000e-8
false
["LlamaForCausalLM"]
false
false
true
false
1
0
false
0
true
1800
[]
false
0
false
false
false
false
false
0
2
0
-
-
no
1-20
of 398