PEFT
Safetensors
dpo
reward-model
preference-learning
lora
trl
interpretability

Commit History

Remove stray Jupyter checkpoint artifact
7a75a30
verified

Shamima commited on

Remove stray Jupyter checkpoint artifact
f007cd7
verified

Shamima commited on

Remove stray Jupyter checkpoint artifact
16d5f1d
verified

Shamima commited on

Remove stray Jupyter checkpoint artifact
7e077ba
verified

Shamima commited on

Remove stray Jupyter checkpoint artifact
06dd8e4
verified

Shamima commited on

Remove stray Jupyter checkpoint artifact
34a9554
verified

Shamima commited on

Remove stray Jupyter checkpoint artifact
d12f116
verified

Shamima commited on

Sync figures
c624257
verified

Shamima commited on

Sync results (5-seed, bias drift, curves)
1fd2e32
verified

Shamima commited on

Document all checkpoints and how to load them
b3892d3
verified

Shamima commited on

Add 4-epoch convergence run: rm_8k_s1 (+8 step checkpoints)
297e07d
verified

Shamima commited on

Add 4-epoch convergence run: rm_8k_s0 (+8 step checkpoints)
59f9949
verified

Shamima commited on

Add 4-epoch convergence run: dpo_beta0.1_8k_s1 (+8 step checkpoints)
a4057f0
verified

Shamima commited on

Add 4-epoch convergence run: dpo_beta0.1_8k_s0 (+8 step checkpoints)
9c6487d
verified

Shamima commited on

Add result figures
523c8b3
verified

Shamima commited on

Add 3-seed eval results
69e6c99
verified

Shamima commited on

Add rm_8k_s2 (3-seed run)
567655b
verified

Shamima commited on

Add dpo_beta0.1_8k_s2 (3-seed run)
c3de32b
verified

Shamima commited on

Add rm_8k_s1 (3-seed run)
486121e
verified

Shamima commited on

Add dpo_beta0.1_8k_s1 (3-seed run)
e480f3f
verified

Shamima commited on

Add rm_8k_s0 (3-seed run)
4f4d788
verified

Shamima commited on

Add dpo_beta0.1_8k_s0 (3-seed run)
b37d819
verified

Shamima commited on

Add rm_8k
7bf361e
verified

Shamima commited on

Add dpo_beta0.1_8k
ed086dd
verified

Shamima commited on

Add sft
69b2dec
verified

Shamima commited on

Add sft_merged
e4781c7
verified

Shamima commited on

Upload README.md with huggingface_hub
c0d918d
verified

Shamima commited on

initial commit
2c14531
verified

Shamima commited on