Upload subspace_ar_bsf071226_v11nl_1x_comp/run_config.yaml with huggingface_hub
Browse files
subspace_ar_bsf071226_v11nl_1x_comp/run_config.yaml
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
ar_cosmag: false
|
| 2 |
+
ar_cosmag_lambda: 0.3
|
| 3 |
+
ar_num_layers: 25
|
| 4 |
+
ar_raw_target: false
|
| 5 |
+
ar_relmse: false
|
| 6 |
+
attn_implementation: sdpa
|
| 7 |
+
av_ortho_penalty: 0.0
|
| 8 |
+
base_ckpt: Qwen/Qwen3-8B
|
| 9 |
+
batch_size: 12
|
| 10 |
+
bullet_min_chars: 1
|
| 11 |
+
comp_mean_bias_npy: /workspace/asherps/nla_analysis/ckpt_snap/free8_minloo075_s900/comp_mean_bias.npy
|
| 12 |
+
compositional: true
|
| 13 |
+
device_map: single
|
| 14 |
+
eval_parquet: null
|
| 15 |
+
eval_skip_rows: 35000
|
| 16 |
+
fixed_bullets: false
|
| 17 |
+
freeze_backbone: false
|
| 18 |
+
full_ft_dtype: fp32
|
| 19 |
+
gradient_accumulation_steps: 1
|
| 20 |
+
gradient_checkpointing: true
|
| 21 |
+
heldout_every: 100
|
| 22 |
+
heldout_parquet: null
|
| 23 |
+
heldout_rows: 1000
|
| 24 |
+
hf_upload: true
|
| 25 |
+
hf_upload_min_steps: 1000
|
| 26 |
+
hf_upload_private: false
|
| 27 |
+
hf_upload_repo: null
|
| 28 |
+
judge_concurrency: 32
|
| 29 |
+
judge_key_env: ANTHROPIC_API_KEY_FALLBACK
|
| 30 |
+
lora_alpha: 16
|
| 31 |
+
lora_r: 128
|
| 32 |
+
lr: 2.0e-05
|
| 33 |
+
lr_warmup_steps: 50
|
| 34 |
+
max_gpu_mem: 0
|
| 35 |
+
max_grad_norm: 1.0
|
| 36 |
+
max_len: 1024
|
| 37 |
+
max_rows: null
|
| 38 |
+
min_lr: 2.0e-06
|
| 39 |
+
mode: ar
|
| 40 |
+
n_samples: 4
|
| 41 |
+
no_wandb: false
|
| 42 |
+
num_bullets: 13
|
| 43 |
+
num_steps: 1298
|
| 44 |
+
paged_optim: false
|
| 45 |
+
parquet: /workspace/asherps/nla-data/subspace_bsf071226_halfchat_v11nl_1x_grounded/ar_sft.parquet
|
| 46 |
+
quant: none
|
| 47 |
+
reward_subspace_npz: null
|
| 48 |
+
sample_every: 2000
|
| 49 |
+
sample_max_new_tokens: 256
|
| 50 |
+
save_dir: /workspace/asherps/nla-ckpts/subspace_ar_bsf071226_v11nl_1x_comp
|
| 51 |
+
save_every: 2000
|
| 52 |
+
seed: 0
|
| 53 |
+
sft_eval_every: 250
|
| 54 |
+
sft_eval_n: 24
|
| 55 |
+
sft_evals: ''
|
| 56 |
+
sidecar: /workspace/asherps/nla-data/subspace_bsf071226_halfchat_v11nl_1x_grounded/ar_sft.parquet
|
| 57 |
+
strip_final_norm: true
|
| 58 |
+
think_column: null
|
| 59 |
+
think_unmasked: false
|
| 60 |
+
uniform_l2_penalty: 0.0
|
| 61 |
+
use_lora: false
|
| 62 |
+
wandb_group: warmstart
|
| 63 |
+
wandb_name: subspace_ar_bsf071226_v11nl_1x_comp
|
| 64 |
+
wandb_project: nla
|
| 65 |
+
wandb_tags: subspace_nla,bsf,v11nl_1x,comp
|