artifacts sync (3055 files)

hs2gemma/exp/rmpool_mbghc1/train/dpo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/htmnpo_coherence/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/htmnpo_conciseness/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/htmnpo_correctness/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/htmnpo_helpfulness/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/inpo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/ipo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/ks/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/maxmin/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/nbs/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/simpo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbghc1/train/unif/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbghc1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/dpo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/htmnpo_coherence/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/htmnpo_conciseness/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/htmnpo_correctness/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/htmnpo_helpfulness/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_htobj: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/inpo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/ipo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/ks/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/maxmin/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/nbs/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/simpo/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre_pref: 1.0
 dataset_splits:

hs2gemma/exp/rmpool_mbgsp1/train/unif/config.yaml

@@ -1,5 +1,5 @@
 model_name_or_path: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
-attn_implementation: eager
+attn_implementation: sdpa
 dataset_mixer:
   /NHNHOME/AIPR/sjkim/gtpref/hs2gemma/exp/rmpool_mbgsp1/pre: 1.0
 dataset_splits:

queue_gm3/ghc.log

@@ -1,2 +1,27 @@
 ===== MB round-1 task=hs2gemma anchor=/NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/ weights={"helpfulness":0.2500,"correctness":0.2500,"coherence":0.2500,"conciseness":0.2500} =====
 allresp built
+scores combined {'pol_42': 2500, 'pol_43': 2500, 'base_44': 2500, 'base_45': 2500}
+{"surplus(2P-1)": {"helpfulness": -0.028, "correctness": -0.027, "coherence": -0.019, "conciseness": 0.002}, "P(pi>mu)": {"helpfulness": 0.486, "correctness": 0.486, "coherence": 0.491, "conciseness": 0.501}, "w_nbs": {"helpfulness": 0.333, "correctness": 0.333, "coherence": 0.333, "conciseness": 0.0}, "w_ks": {"helpfulness": 0.252, "correctness": 0.252, "coherence": 0.252, "conciseness": 0.245}, "worst": "helpfulness", "n": 5000}
+RMPOOL_hs2gemma_DONE
+[build_bon] 2500 pairs (train 2250/test 250) weights {'helpfulness': 0.25, 'correctness': 0.25, 'coherence': 0.25, 'conciseness': 0.25}
+[prep hs2gemma EXP=exp/rmpool_mbghc1] objs: helpfulness correctness coherence conciseness parent: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 80759.78 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 80337.66 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 21055.32 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 20897.54 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 90015.11 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 89707.07 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 24861.91 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 24662.51 examples/s]
+[prep] pre_pref + pre_htobj built
+dpo htmnpo_coherence htmnpo_conciseness htmnpo_correctness htmnpo_helpfulness inpo ipo ks maxmin nbs simpo unif 
+/NHNHOME/AIPR/sjkim/gtpref/train_gpu.sh: line 15: 962667 Killed                  CUDA_VISIBLE_DEVICES=${GA[$((g%NGA))]} $V -m accelerate.commands.launch --config_file $P/accelerate_configs/single_gpu.yaml --num_processes=1 --main_process_port=$((PB+g)) -m mnpo_scripts.run_mnpo $OUT/config.yaml > $OUT/train.log 2>&1
+TRAIN_DONE aiprlab-polymer

queue_gm3/gks.log

@@ -26,3 +26,30 @@ dpo htmnpo_coherence htmnpo_conciseness htmnpo_correctness htmnpo_helpfulness in
 TRAIN_DONE aiprlab-polymer
 ===== MB round-1 task=hs2gemma anchor=/NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/ weights={"helpfulness": 0.25, "correctness": 0.25, "coherence": 0.25, "conciseness": 0.25} =====
 allresp built
+scores combined {'pol_42': 2500, 'pol_43': 2500, 'base_44': 2500, 'base_45': 2500}
+{"surplus(2P-1)": {"helpfulness": -0.028, "correctness": -0.027, "coherence": -0.019, "conciseness": 0.002}, "P(pi>mu)": {"helpfulness": 0.486, "correctness": 0.486, "coherence": 0.491, "conciseness": 0.501}, "w_nbs": {"helpfulness": 0.333, "correctness": 0.333, "coherence": 0.333, "conciseness": 0.0}, "w_ks": {"helpfulness": 0.252, "correctness": 0.252, "coherence": 0.252, "conciseness": 0.245}, "worst": "helpfulness", "n": 5000}
+RMPOOL_hs2gemma_DONE
+[build_bon] 2500 pairs (train 2250/test 250) weights {'helpfulness': 0.25, 'correctness': 0.25, 'coherence': 0.25, 'conciseness': 0.25}
+[prep hs2gemma EXP=exp/rmpool_mbgks1] objs: helpfulness correctness coherence conciseness parent: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 89809.52 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 89321.60 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 23495.92 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 23299.62 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 96440.49 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 96085.04 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 24603.49 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 24407.63 examples/s]
+[prep] pre_pref + pre_htobj built
+dpo htmnpo_coherence htmnpo_conciseness htmnpo_correctness htmnpo_helpfulness inpo ipo ks maxmin nbs simpo unif 
+/NHNHOME/AIPR/sjkim/gtpref/train_gpu.sh: line 15: 963572 Killed                  CUDA_VISIBLE_DEVICES=${GA[$((g%NGA))]} $V -m accelerate.commands.launch --config_file $P/accelerate_configs/single_gpu.yaml --num_processes=1 --main_process_port=$((PB+g)) -m mnpo_scripts.run_mnpo $OUT/config.yaml > $OUT/train.log 2>&1
+TRAIN_DONE aiprlab-polymer
+===== MB round-1 task=hs2gemma anchor=/NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/ weights={"helpfulness": 0.25, "correctness": 0.25, "coherence": 0.25, "conciseness": 0.25} =====
+allresp built

queue_gm3/gnb.log

@@ -26,3 +26,30 @@ dpo htmnpo_coherence htmnpo_conciseness htmnpo_correctness htmnpo_helpfulness in
 TRAIN_DONE aiprlab-polymer
 ===== MB round-1 task=hs2gemma anchor=/NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/ weights={"helpfulness": 0.25, "correctness": 0.25, "coherence": 0.25, "conciseness": 0.25} =====
 allresp built
+scores combined {'pol_42': 2500, 'pol_43': 2500, 'base_44': 2500, 'base_45': 2500}
+{"surplus(2P-1)": {"helpfulness": -0.028, "correctness": -0.027, "coherence": -0.019, "conciseness": 0.002}, "P(pi>mu)": {"helpfulness": 0.486, "correctness": 0.486, "coherence": 0.491, "conciseness": 0.501}, "w_nbs": {"helpfulness": 0.333, "correctness": 0.333, "coherence": 0.333, "conciseness": 0.0}, "w_ks": {"helpfulness": 0.252, "correctness": 0.252, "coherence": 0.252, "conciseness": 0.245}, "worst": "helpfulness", "n": 5000}
+RMPOOL_hs2gemma_DONE
+[build_bon] 2500 pairs (train 2250/test 250) weights {'helpfulness': 0.25, 'correctness': 0.25, 'coherence': 0.25, 'conciseness': 0.25}
+[prep hs2gemma EXP=exp/rmpool_mbgnb1] objs: helpfulness correctness coherence conciseness parent: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 92844.55 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 92324.09 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 23247.44 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 23047.65 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 96304.67 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 95938.52 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 24777.90 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 24574.08 examples/s]
+[prep] pre_pref + pre_htobj built
+dpo htmnpo_coherence htmnpo_conciseness htmnpo_correctness htmnpo_helpfulness inpo ipo ks maxmin nbs simpo unif 
+/NHNHOME/AIPR/sjkim/gtpref/train_gpu.sh: line 15: 963757 Killed                  CUDA_VISIBLE_DEVICES=${GA[$((g%NGA))]} $V -m accelerate.commands.launch --config_file $P/accelerate_configs/single_gpu.yaml --num_processes=1 --main_process_port=$((PB+g)) -m mnpo_scripts.run_mnpo $OUT/config.yaml > $OUT/train.log 2>&1
+TRAIN_DONE aiprlab-polymer
+===== MB round-1 task=hs2gemma anchor=/NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/ weights={"helpfulness": 0.25, "correctness": 0.25, "coherence": 0.25, "conciseness": 0.25} =====
+allresp built

queue_gm3/gsp.log

@@ -1,2 +1,27 @@
 ===== MB round-1 task=hs2gemma anchor=/NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/ weights={"helpfulness":0.2500,"correctness":0.2500,"coherence":0.2500,"conciseness":0.2500} =====
 allresp built
+scores combined {'pol_42': 2500, 'pol_43': 2500, 'base_44': 2500, 'base_45': 2500}
+{"surplus(2P-1)": {"helpfulness": -0.035, "correctness": -0.033, "coherence": -0.028, "conciseness": 0.001}, "P(pi>mu)": {"helpfulness": 0.482, "correctness": 0.484, "coherence": 0.486, "conciseness": 0.5}, "w_nbs": {"helpfulness": 0.333, "correctness": 0.333, "coherence": 0.333, "conciseness": 0.001}, "w_ks": {"helpfulness": 0.251, "correctness": 0.251, "coherence": 0.251, "conciseness": 0.247}, "worst": "helpfulness", "n": 5000}
+RMPOOL_hs2gemma_DONE
+[build_bon] 2500 pairs (train 2250/test 250) weights {'helpfulness': 0.25, 'correctness': 0.25, 'coherence': 0.25, 'conciseness': 0.25}
+[prep hs2gemma EXP=exp/rmpool_mbgsp1] objs: helpfulness correctness coherence conciseness parent: /NHNHOME/AIPR/sjkim/policy_bases/models--google--gemma-2-9b-it/snapshots/11c9b309abf73637e4b6f9a3fa1e92e615547819/
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 90541.05 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 90029.71 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 19990.01 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 19844.74 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/2250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 105152.03 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 2250/2250 [00:00<00:00, 104726.11 examples/s]
+
+Saving the dataset (0/1 shards):   0%|          | 0/250 [00:00<?, ? examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 25249.25 examples/s]
+Saving the dataset (1/1 shards): 100%|██████████| 250/250 [00:00<00:00, 25034.04 examples/s]
+[prep] pre_pref + pre_htobj built
+dpo htmnpo_coherence htmnpo_conciseness htmnpo_correctness htmnpo_helpfulness inpo ipo ks maxmin nbs simpo unif 
+/NHNHOME/AIPR/sjkim/gtpref/train_gpu.sh: line 15: 963363 Killed                  CUDA_VISIBLE_DEVICES=${GA[$((g%NGA))]} $V -m accelerate.commands.launch --config_file $P/accelerate_configs/single_gpu.yaml --num_processes=1 --main_process_port=$((PB+g)) -m mnpo_scripts.run_mnpo $OUT/config.yaml > $OUT/train.log 2>&1
+TRAIN_DONE aiprlab-polymer

queue_gm3/queue.log

@@ -6,3 +6,6 @@
 [08-07 19:04] chain gnb arm=nbs gpu=1 rounds='1 2 3 4 5 6 7 '
 [08-07 19:04] chain gsp arm=simpo gpu=3 rounds='1 2 3 4 5 6 7 '
 [08-07 19:04] chain ghc arm=htmnpo_coherence gpu=2 rounds='1 2 3 4 5 6 7 '
+[08-07 19:20] ===== gemma panel on GPUs 0 1 =====
+[08-07 19:20] chain gks arm=ks gpu=1 rounds='1 2 3 4 5 6 7 '
+[08-07 19:20] chain gnb arm=nbs gpu=0 rounds='1 2 3 4 5 6 7 '

queue_ipo3/queue.log

@@ -0,0 +1 @@
+[08-07 19:22] ===== train IPO at seeds 43 and 44 (GPUs 2 and 3) =====

queue_ipo3/train_p43.log

@@ -0,0 +1,2 @@
+skip helpsteer2:ipo
+TRAIN_DONE aiprlab-polymer