Alexander-vishnevskiy's workspace
Runs
54
Name
54 visualized
State
Notes
User
Tags
Created
Runtime
Sweep
_name_or_path
activation_checkpointing
activation_function
adafactor
adam_beta1
adam_beta2
adam_epsilon
add_cross_attention
architectures
attention_dropout
attn_pdrop
auto_find_batch_size
auto_map.AutoConfig
auto_map.AutoModel
auto_map.AutoModelForCausalLM
auto_map.AutoModelForQuestionAnswering
auto_map.AutoModelForSequenceClassification
auto_map.AutoModelForTokenClassification
batch_size
bf16
bf16_full_eval
bos_token_id
chunk_size_feed_forward
classifier_dropout
dataloader_drop_last
dataloader_num_workers
dataloader_pin_memory
datasets
ddp_timeout
debug
disable_tqdm
diversity_penalty
do_eval
do_first_eval
do_predict
do_sample
do_train
early_stopping
embd_pdrop
encoder_no_repeat_ngram_size
eos_token_id
eval_batch_size
eval_delay
eval_every
Finished
alexander-vishnevskiy
13h 11m 3s
-
/data/avishnevskiy/experiments/sft-btlm
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Killed
alexander-vishnevskiy
4h 10m 57s
-
/data/avishnevskiy/experiments/btlm-sft-alpaca-hh
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Killed
alexander-vishnevskiy
9h 52m 12s
-
/data/avishnevskiy/experiments/btlm-sft-alpaca-hh
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Finished
alexander-vishnevskiy
3h 12m 47s
-
/data/avishnevskiy/experiments/btlm-alpaca-sft
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Finished
alexander-vishnevskiy
6h 59m 52s
-
-
false
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
64
-
-
-
-
-
-
-
-
["alpaca","hh"]
-
false
-
-
-
true
-
-
-
-
-
-
-
1
-
19968
Finished
alexander-vishnevskiy
11h 20m 27s
-
lomahony/eleuther-pythia2.8b-hh-sft
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
2h 52m 42s
-
/data/avishnevskiy/experiments/btlm-alpaca-sft
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Finished
alexander-vishnevskiy
12h 19m 26s
-
/data/avishnevskiy/experiments/pythia2.8_epoch2-20231120-041550/LATEST
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
12h 19m 5s
-
/data/avishnevskiy/experiments/trl_dpo_pythia_fp16-20231011-010059/LATEST
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
23h 18m 53s
-
/data/avishnevskiy/experiments/pythia410-sft
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Killed
alexander-vishnevskiy
3h 55m 21s
-
/data/avishnevskiy/experiments/pythia410-sft
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
1h 12m 50s
-
-
false
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
64
-
-
-
-
-
-
-
-
["hh"]
-
false
-
-
-
true
-
-
-
-
-
-
-
1
-
19968
Finished
alexander-vishnevskiy
9h 3m 40s
-
/data/avishnevskiy/experiments/dpo_pythia1.4-20231114-062744/LATEST
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
12h 23m 32s
-
/data/avishnevskiy/experiments/sft-btlm
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Finished
alexander-vishnevskiy
12h 41m 10s
-
/data/avishnevskiy/experiments/btlm-alpaca-sft
-
swiglu
false
0.9
0.999
1.0000e-8
false
["BTLMLMHeadModel"]
-
0
false
cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForQuestionAnswering
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForSequenceClassification
cerebras/btlm-3b-8k-base--modeling_btlm.BTLMForTokenClassification
-
true
false
50256
0
-
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
0
0
50256
-
0
-
Finished
alexander-vishnevskiy
11h 14m 54s
-
/data/avishnevskiy/experiments/pythia1.4-sft
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Killed
alexander-vishnevskiy
9h 20m 30s
-
/data/avishnevskiy/experiments/pythia1.4-sft
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
1h 43m 36s
-
-
false
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
64
-
-
-
-
-
-
-
-
["alpaca"]
-
false
-
-
-
true
-
-
-
-
-
-
-
1
-
19968
Finished
alexander-vishnevskiy
8h 19m 51s
-
/data/avishnevskiy/experiments/pythia1.4-sft
-
-
false
0.9
0.999
1.0000e-8
false
["GPTNeoXForCausalLM"]
0
-
false
-
-
-
-
-
-
-
true
false
0
0
0.1
false
0
true
-
1800
[]
false
0
true
-
false
false
false
false
-
0
0
-
0
-
Finished
alexander-vishnevskiy
2h 53m 14s
-
-
false
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
64
-
-
-
-
-
-
-
-
["hh"]
-
false
-
-
-
true
-
-
-
-
-
-
-
1
-
59968
1-20
of 54