infinite_prompt_pipeline v. main

infinite_prompt_pipeline @a50cc07/Adding `infinite_dataloader` since previous solution doesn't seem to work as expected/2023-03-07 main @ded2e5e/fix(ppo_trainer): update `AdaptiveKLController` with correct KL (#361)/2023-03-13

Sorry

Created on March 13|Last edited on March 13

Comment

﻿
ilql_randomwalks/GPT2Config/1gpu﻿
metrics/lengths@beta=100
metrics/lengths@beta=100
050100150Step20406080100
metrics/optimality@beta=1
metrics/optimality@beta=1
050100150Step00.10.20.30.4
metrics/lengths@beta=1
metrics/lengths@beta=1
050100150Step5060708090100
Run set2
﻿
﻿
﻿
Run set2
﻿
ppo_sentiments/gpt2-imdb/1gpu﻿
Run set2
﻿
﻿
﻿
Run set2
﻿
ppo_randomwalks/randomwalks/1gpu﻿
Run set2
﻿
﻿
﻿
Run set2
﻿
sft_sentiments/gpt2/1gpu﻿
Run set2
﻿
﻿
﻿
Run set2
﻿
ilql_sentiments/gpt2/1gpu﻿
Run set2
﻿
﻿
﻿
Run set2
﻿
ppo_hh/pythia-6B-static-sft/7gpus﻿
Run set2
﻿
﻿
﻿
Run set2
﻿
﻿

Add a comment