Hugging Face's logo

radna
/
DeepSeek-R1-14B-RL-70-RLS3-ckpts

Safetensors
Model card Files
xet
Community
radna commited on
Commit
d1d788c
·
verified ·
1 Parent(s): c6368e6

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. checkpoint-10/adapter_config.json +4 -4
  2. checkpoint-10/trainer_state.json +11 -11
  3. checkpoint-10/training_args.bin +1 -1
  4. checkpoint-12/adapter_config.json +4 -4
  5. checkpoint-12/trainer_state.json +14 -14
  6. checkpoint-12/training_args.bin +1 -1
  7. checkpoint-14/adapter_config.json +4 -4
  8. checkpoint-14/trainer_state.json +16 -16
  9. checkpoint-14/training_args.bin +1 -1
  10. checkpoint-16/adapter_config.json +4 -4
  11. checkpoint-16/trainer_state.json +18 -18
  12. checkpoint-16/training_args.bin +1 -1
  13. checkpoint-18/adapter_config.json +4 -4
  14. checkpoint-18/trainer_state.json +20 -20
  15. checkpoint-18/training_args.bin +1 -1
  16. checkpoint-2/adapter_config.json +4 -4
  17. checkpoint-2/trainer_state.json +2 -2
  18. checkpoint-2/training_args.bin +1 -1
  19. checkpoint-20/adapter_config.json +4 -4
  20. checkpoint-20/trainer_state.json +22 -22
  21. checkpoint-20/training_args.bin +1 -1
  22. checkpoint-22/adapter_config.json +4 -4
  23. checkpoint-22/trainer_state.json +24 -24
  24. checkpoint-22/training_args.bin +1 -1
  25. checkpoint-24/adapter_config.json +4 -4
  26. checkpoint-24/trainer_state.json +27 -27
  27. checkpoint-24/training_args.bin +1 -1
  28. checkpoint-26/adapter_config.json +4 -4
  29. checkpoint-26/trainer_state.json +29 -29
  30. checkpoint-26/training_args.bin +1 -1
  31. checkpoint-28/adapter_config.json +4 -4
  32. checkpoint-28/trainer_state.json +31 -31
  33. checkpoint-28/training_args.bin +1 -1
  34. checkpoint-30/README.md +202 -0
  35. checkpoint-30/adapter_config.json +37 -0
  36. checkpoint-30/adapter_model.safetensors +3 -0
  37. checkpoint-30/additional_config.json +1 -0
  38. checkpoint-30/global_step30/bf16_zero_pp_rank_16_mp_rank_00_optim_states.pt +3 -0
  39. checkpoint-30/global_step30/bf16_zero_pp_rank_17_mp_rank_00_optim_states.pt +3 -0
  40. checkpoint-30/global_step30/bf16_zero_pp_rank_18_mp_rank_00_optim_states.pt +3 -0
  41. checkpoint-30/global_step30/bf16_zero_pp_rank_19_mp_rank_00_optim_states.pt +3 -0
  42. checkpoint-30/global_step30/bf16_zero_pp_rank_20_mp_rank_00_optim_states.pt +3 -0
  43. checkpoint-30/global_step30/bf16_zero_pp_rank_21_mp_rank_00_optim_states.pt +3 -0
  44. checkpoint-30/global_step30/bf16_zero_pp_rank_22_mp_rank_00_optim_states.pt +3 -0
  45. checkpoint-30/global_step30/bf16_zero_pp_rank_23_mp_rank_00_optim_states.pt +3 -0
  46. checkpoint-30/global_step30/mp_rank_00_model_states.pt +3 -0
  47. checkpoint-30/latest +1 -0
  48. checkpoint-30/rng_state_16.pth +3 -0
  49. checkpoint-30/rng_state_17.pth +3 -0
  50. checkpoint-30/rng_state_18.pth +3 -0
checkpoint-10/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-10/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  }
checkpoint-10/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-12/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-12/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
checkpoint-12/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-14/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-14/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  }
checkpoint-14/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-16/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-16/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  }
checkpoint-16/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-18/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-18/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
@@ -269,7 +269,7 @@
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
- "memory(GiB)": 180.29,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
@@ -283,7 +283,7 @@
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
- "memory(GiB)": 180.29,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
 
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
+ "memory(GiB)": 187.02,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
 
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
+ "memory(GiB)": 187.02,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
checkpoint-18/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-2/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-2/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  }
checkpoint-2/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-20/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-20/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
@@ -269,7 +269,7 @@
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
- "memory(GiB)": 180.29,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
@@ -283,7 +283,7 @@
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
- "memory(GiB)": 180.29,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
@@ -311,7 +311,7 @@
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
- "memory(GiB)": 180.29,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
@@ -327,7 +327,7 @@
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
- "memory(GiB)": 180.29,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
 
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
+ "memory(GiB)": 187.02,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
 
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
+ "memory(GiB)": 187.02,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
 
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
+ "memory(GiB)": 187.02,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
 
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
+ "memory(GiB)": 187.02,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  }
checkpoint-20/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-22/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-22/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
@@ -269,7 +269,7 @@
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
- "memory(GiB)": 180.29,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
@@ -283,7 +283,7 @@
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
- "memory(GiB)": 180.29,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
@@ -311,7 +311,7 @@
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
- "memory(GiB)": 180.29,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
@@ -327,7 +327,7 @@
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
- "memory(GiB)": 180.29,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
@@ -339,7 +339,7 @@
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
- "memory(GiB)": 180.29,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
@@ -355,7 +355,7 @@
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
- "memory(GiB)": 180.29,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
 
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
+ "memory(GiB)": 187.02,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
 
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
+ "memory(GiB)": 187.02,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
 
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
+ "memory(GiB)": 187.02,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
 
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
+ "memory(GiB)": 187.02,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
 
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
+ "memory(GiB)": 187.02,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
 
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
+ "memory(GiB)": 187.02,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  }
checkpoint-22/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-24/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-24/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
@@ -269,7 +269,7 @@
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
- "memory(GiB)": 180.29,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
@@ -283,7 +283,7 @@
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
- "memory(GiB)": 180.29,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
@@ -311,7 +311,7 @@
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
- "memory(GiB)": 180.29,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
@@ -327,7 +327,7 @@
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
- "memory(GiB)": 180.29,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
@@ -339,7 +339,7 @@
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
- "memory(GiB)": 180.29,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
@@ -355,7 +355,7 @@
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
- "memory(GiB)": 180.29,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  },
@@ -367,7 +367,7 @@
367
  "kl": 0.1796875,
368
  "learning_rate": 7.74754489035403e-05,
369
  "loss": -0.03423420712351799,
370
- "memory(GiB)": 180.29,
371
  "response_clip_ratio": 0.583984375,
372
  "reward": 0.034468831261619925,
373
  "reward_std": 0.11841745302081108,
@@ -381,7 +381,7 @@
381
  "grad_norm": 0.014131724834442139,
382
  "learning_rate": 7.500000000000001e-05,
383
  "loss": -0.03426633030176163,
384
- "memory(GiB)": 180.29,
385
  "step": 24,
386
  "train_speed(iter/s)": 0.000427
387
  },
@@ -396,7 +396,7 @@
396
  "eval_reward_std": 0.10456253588199615,
397
  "eval_rewards/CosineReward": 0.04339282959699631,
398
  "eval_rewards/RepetitionPenalty": 0.0,
399
- "eval_runtime": 1045.0616,
400
  "eval_samples_per_second": 0.001,
401
  "eval_steps_per_second": 0.001,
402
  "step": 24
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
 
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
+ "memory(GiB)": 187.02,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
 
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
+ "memory(GiB)": 187.02,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
 
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
+ "memory(GiB)": 187.02,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
 
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
+ "memory(GiB)": 187.02,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
 
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
+ "memory(GiB)": 187.02,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
 
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
+ "memory(GiB)": 187.02,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  },
 
367
  "kl": 0.1796875,
368
  "learning_rate": 7.74754489035403e-05,
369
  "loss": -0.03423420712351799,
370
+ "memory(GiB)": 187.02,
371
  "response_clip_ratio": 0.583984375,
372
  "reward": 0.034468831261619925,
373
  "reward_std": 0.11841745302081108,
 
381
  "grad_norm": 0.014131724834442139,
382
  "learning_rate": 7.500000000000001e-05,
383
  "loss": -0.03426633030176163,
384
+ "memory(GiB)": 187.02,
385
  "step": 24,
386
  "train_speed(iter/s)": 0.000427
387
  },
 
396
  "eval_reward_std": 0.10456253588199615,
397
  "eval_rewards/CosineReward": 0.04339282959699631,
398
  "eval_rewards/RepetitionPenalty": 0.0,
399
+ "eval_runtime": 1045.0642,
400
  "eval_samples_per_second": 0.001,
401
  "eval_steps_per_second": 0.001,
402
  "step": 24
checkpoint-24/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-26/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-26/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
@@ -269,7 +269,7 @@
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
- "memory(GiB)": 180.29,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
@@ -283,7 +283,7 @@
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
- "memory(GiB)": 180.29,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
@@ -311,7 +311,7 @@
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
- "memory(GiB)": 180.29,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
@@ -327,7 +327,7 @@
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
- "memory(GiB)": 180.29,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
@@ -339,7 +339,7 @@
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
- "memory(GiB)": 180.29,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
@@ -355,7 +355,7 @@
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
- "memory(GiB)": 180.29,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  },
@@ -367,7 +367,7 @@
367
  "kl": 0.1796875,
368
  "learning_rate": 7.74754489035403e-05,
369
  "loss": -0.03423420712351799,
370
- "memory(GiB)": 180.29,
371
  "response_clip_ratio": 0.583984375,
372
  "reward": 0.034468831261619925,
373
  "reward_std": 0.11841745302081108,
@@ -381,7 +381,7 @@
381
  "grad_norm": 0.014131724834442139,
382
  "learning_rate": 7.500000000000001e-05,
383
  "loss": -0.03426633030176163,
384
- "memory(GiB)": 180.29,
385
  "step": 24,
386
  "train_speed(iter/s)": 0.000427
387
  },
@@ -396,7 +396,7 @@
396
  "eval_reward_std": 0.10456253588199615,
397
  "eval_rewards/CosineReward": 0.04339282959699631,
398
  "eval_rewards/RepetitionPenalty": 0.0,
399
- "eval_runtime": 1045.0616,
400
  "eval_samples_per_second": 0.001,
401
  "eval_steps_per_second": 0.001,
402
  "step": 24
@@ -409,7 +409,7 @@
409
  "kl": 0.1800537109375,
410
  "learning_rate": 7.243995901002312e-05,
411
  "loss": -0.02097315341234207,
412
- "memory(GiB)": 180.29,
413
  "response_clip_ratio": 0.6171875,
414
  "reward": 0.03010205877944827,
415
  "reward_std": 0.10742511600255966,
@@ -425,7 +425,7 @@
425
  "kl": 0.18408203125,
426
  "learning_rate": 6.980398830195785e-05,
427
  "loss": -0.02103913575410843,
428
- "memory(GiB)": 180.29,
429
  "step": 26,
430
  "train_speed(iter/s)": 0.000421
431
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
 
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
+ "memory(GiB)": 187.02,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
 
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
+ "memory(GiB)": 187.02,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
 
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
+ "memory(GiB)": 187.02,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
 
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
+ "memory(GiB)": 187.02,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
 
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
+ "memory(GiB)": 187.02,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
 
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
+ "memory(GiB)": 187.02,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  },
 
367
  "kl": 0.1796875,
368
  "learning_rate": 7.74754489035403e-05,
369
  "loss": -0.03423420712351799,
370
+ "memory(GiB)": 187.02,
371
  "response_clip_ratio": 0.583984375,
372
  "reward": 0.034468831261619925,
373
  "reward_std": 0.11841745302081108,
 
381
  "grad_norm": 0.014131724834442139,
382
  "learning_rate": 7.500000000000001e-05,
383
  "loss": -0.03426633030176163,
384
+ "memory(GiB)": 187.02,
385
  "step": 24,
386
  "train_speed(iter/s)": 0.000427
387
  },
 
396
  "eval_reward_std": 0.10456253588199615,
397
  "eval_rewards/CosineReward": 0.04339282959699631,
398
  "eval_rewards/RepetitionPenalty": 0.0,
399
+ "eval_runtime": 1045.0642,
400
  "eval_samples_per_second": 0.001,
401
  "eval_steps_per_second": 0.001,
402
  "step": 24
 
409
  "kl": 0.1800537109375,
410
  "learning_rate": 7.243995901002312e-05,
411
  "loss": -0.02097315341234207,
412
+ "memory(GiB)": 187.02,
413
  "response_clip_ratio": 0.6171875,
414
  "reward": 0.03010205877944827,
415
  "reward_std": 0.10742511600255966,
 
425
  "kl": 0.18408203125,
426
  "learning_rate": 6.980398830195785e-05,
427
  "loss": -0.02103913575410843,
428
+ "memory(GiB)": 187.02,
429
  "step": 26,
430
  "train_speed(iter/s)": 0.000421
431
  }
checkpoint-26/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-28/adapter_config.json CHANGED
@@ -23,13 +23,13 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "down_proj",
27
- "o_proj",
28
- "q_proj",
29
  "v_proj",
30
  "up_proj",
 
31
  "gate_proj",
32
- "k_proj"
 
 
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
 
 
 
26
  "v_proj",
27
  "up_proj",
28
+ "k_proj",
29
  "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
35
  "use_dora": false,
checkpoint-28/trainer_state.json CHANGED
@@ -17,7 +17,7 @@
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
- "memory(GiB)": 180.29,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
@@ -33,7 +33,7 @@
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
- "memory(GiB)": 180.29,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
@@ -45,7 +45,7 @@
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
- "memory(GiB)": 180.29,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
@@ -61,7 +61,7 @@
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
- "memory(GiB)": 180.29,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
@@ -73,7 +73,7 @@
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
- "memory(GiB)": 180.29,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
@@ -87,7 +87,7 @@
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
- "memory(GiB)": 180.29,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
@@ -102,7 +102,7 @@
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
- "eval_runtime": 1030.1223,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
@@ -115,7 +115,7 @@
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
- "memory(GiB)": 180.29,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
@@ -131,7 +131,7 @@
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
- "memory(GiB)": 180.29,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
@@ -143,7 +143,7 @@
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
- "memory(GiB)": 180.29,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
@@ -159,7 +159,7 @@
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
- "memory(GiB)": 180.29,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
@@ -171,7 +171,7 @@
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
- "memory(GiB)": 180.29,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
@@ -185,7 +185,7 @@
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
- "memory(GiB)": 180.29,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
@@ -200,7 +200,7 @@
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
- "eval_runtime": 1025.9045,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
@@ -213,7 +213,7 @@
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
- "memory(GiB)": 180.29,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
@@ -229,7 +229,7 @@
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
- "memory(GiB)": 180.29,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
@@ -241,7 +241,7 @@
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
- "memory(GiB)": 180.29,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
@@ -257,7 +257,7 @@
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
- "memory(GiB)": 180.29,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
@@ -269,7 +269,7 @@
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
- "memory(GiB)": 180.29,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
@@ -283,7 +283,7 @@
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
- "memory(GiB)": 180.29,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
@@ -311,7 +311,7 @@
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
- "memory(GiB)": 180.29,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
@@ -327,7 +327,7 @@
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
- "memory(GiB)": 180.29,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
@@ -339,7 +339,7 @@
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
- "memory(GiB)": 180.29,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
@@ -355,7 +355,7 @@
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
- "memory(GiB)": 180.29,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  },
@@ -367,7 +367,7 @@
367
  "kl": 0.1796875,
368
  "learning_rate": 7.74754489035403e-05,
369
  "loss": -0.03423420712351799,
370
- "memory(GiB)": 180.29,
371
  "response_clip_ratio": 0.583984375,
372
  "reward": 0.034468831261619925,
373
  "reward_std": 0.11841745302081108,
@@ -381,7 +381,7 @@
381
  "grad_norm": 0.014131724834442139,
382
  "learning_rate": 7.500000000000001e-05,
383
  "loss": -0.03426633030176163,
384
- "memory(GiB)": 180.29,
385
  "step": 24,
386
  "train_speed(iter/s)": 0.000427
387
  },
@@ -396,7 +396,7 @@
396
  "eval_reward_std": 0.10456253588199615,
397
  "eval_rewards/CosineReward": 0.04339282959699631,
398
  "eval_rewards/RepetitionPenalty": 0.0,
399
- "eval_runtime": 1045.0616,
400
  "eval_samples_per_second": 0.001,
401
  "eval_steps_per_second": 0.001,
402
  "step": 24
@@ -409,7 +409,7 @@
409
  "kl": 0.1800537109375,
410
  "learning_rate": 7.243995901002312e-05,
411
  "loss": -0.02097315341234207,
412
- "memory(GiB)": 180.29,
413
  "response_clip_ratio": 0.6171875,
414
  "reward": 0.03010205877944827,
415
  "reward_std": 0.10742511600255966,
@@ -425,7 +425,7 @@
425
  "kl": 0.18408203125,
426
  "learning_rate": 6.980398830195785e-05,
427
  "loss": -0.02103913575410843,
428
- "memory(GiB)": 180.29,
429
  "step": 26,
430
  "train_speed(iter/s)": 0.000421
431
  },
@@ -437,7 +437,7 @@
437
  "kl": 0.174560546875,
438
  "learning_rate": 6.710100716628344e-05,
439
  "loss": -0.03593946248292923,
440
- "memory(GiB)": 180.29,
441
  "response_clip_ratio": 0.513671875,
442
  "reward": 0.04752760287374258,
443
  "reward_std": 0.14935147762298584,
@@ -453,7 +453,7 @@
453
  "kl": 0.182373046875,
454
  "learning_rate": 6.434016163555452e-05,
455
  "loss": -0.03595500811934471,
456
- "memory(GiB)": 180.29,
457
  "step": 28,
458
  "train_speed(iter/s)": 0.000422
459
  }
 
17
  "kl": 0.0,
18
  "learning_rate": 1.6666666666666667e-05,
19
  "loss": -0.11016345024108887,
20
+ "memory(GiB)": 176.98,
21
  "response_clip_ratio": 0.11328125,
22
  "reward": -0.002658387296833098,
23
  "reward_std": 0.06134121119976044,
 
33
  "kl": 0.0,
34
  "learning_rate": 3.3333333333333335e-05,
35
  "loss": -0.11016345024108887,
36
+ "memory(GiB)": 176.98,
37
  "step": 2,
38
  "train_speed(iter/s)": 0.000466
39
  },
 
45
  "kl": 9.50181856751442e-07,
46
  "learning_rate": 5e-05,
47
  "loss": -0.06604708731174469,
48
+ "memory(GiB)": 176.98,
49
  "response_clip_ratio": 0.13671875,
50
  "reward": 0.0006296975770965219,
51
  "reward_std": 0.07172460854053497,
 
61
  "kl": 1.1101365089416504e-05,
62
  "learning_rate": 6.666666666666667e-05,
63
  "loss": -0.06727766245603561,
64
+ "memory(GiB)": 176.98,
65
  "step": 4,
66
  "train_speed(iter/s)": 0.000458
67
  },
 
73
  "kl": 0.00017762184143066406,
74
  "learning_rate": 8.333333333333334e-05,
75
  "loss": -0.09315311908721924,
76
+ "memory(GiB)": 176.98,
77
  "response_clip_ratio": 0.119140625,
78
  "reward": -0.005135859013535082,
79
  "reward_std": 0.07994875870645046,
 
87
  "grad_norm": 0.18263348937034607,
88
  "learning_rate": 0.0001,
89
  "loss": -0.1041698157787323,
90
+ "memory(GiB)": 176.98,
91
  "step": 6,
92
  "train_speed(iter/s)": 0.000458
93
  },
 
102
  "eval_reward_std": 0.08769983053207397,
103
  "eval_rewards/CosineReward": 0.012996694073081017,
104
  "eval_rewards/RepetitionPenalty": 0.0,
105
+ "eval_runtime": 1030.1126,
106
  "eval_samples_per_second": 0.001,
107
  "eval_steps_per_second": 0.001,
108
  "step": 6
 
115
  "kl": 0.017406463623046875,
116
  "learning_rate": 9.991540791356342e-05,
117
  "loss": -0.051375165581703186,
118
+ "memory(GiB)": 176.98,
119
  "response_clip_ratio": 0.1484375,
120
  "reward": 0.004909618757665157,
121
  "reward_std": 0.08167182095348835,
 
131
  "kl": 0.089599609375,
132
  "learning_rate": 9.966191788709716e-05,
133
  "loss": -0.05105742812156677,
134
+ "memory(GiB)": 176.98,
135
  "step": 8,
136
  "train_speed(iter/s)": 0.000433
137
  },
 
143
  "kl": 0.0963134765625,
144
  "learning_rate": 9.924038765061042e-05,
145
  "loss": -0.05842069163918495,
146
+ "memory(GiB)": 176.98,
147
  "response_clip_ratio": 0.255859375,
148
  "reward": 0.03643610421568155,
149
  "reward_std": 0.11898956261575222,
 
159
  "kl": 0.1185302734375,
160
  "learning_rate": 9.865224352899119e-05,
161
  "loss": -0.06491819024085999,
162
+ "memory(GiB)": 176.98,
163
  "step": 10,
164
  "train_speed(iter/s)": 0.000436
165
  },
 
171
  "kl": 0.1275634765625,
172
  "learning_rate": 9.789947561577445e-05,
173
  "loss": -0.04600231721997261,
174
+ "memory(GiB)": 187.02,
175
  "response_clip_ratio": 0.361328125,
176
  "reward": 0.023204635945148766,
177
  "reward_std": 0.10593634657561779,
 
185
  "grad_norm": 0.05781339108943939,
186
  "learning_rate": 9.698463103929542e-05,
187
  "loss": -0.05069056898355484,
188
+ "memory(GiB)": 187.02,
189
  "step": 12,
190
  "train_speed(iter/s)": 0.000439
191
  },
 
200
  "eval_reward_std": 0.10685288906097412,
201
  "eval_rewards/CosineReward": 0.03234308212995529,
202
  "eval_rewards/RepetitionPenalty": 0.0,
203
+ "eval_runtime": 1025.9048,
204
  "eval_samples_per_second": 0.001,
205
  "eval_steps_per_second": 0.001,
206
  "step": 12
 
213
  "kl": 0.151123046875,
214
  "learning_rate": 9.591080534401371e-05,
215
  "loss": -0.02191038429737091,
216
+ "memory(GiB)": 187.02,
217
  "response_clip_ratio": 0.419921875,
218
  "reward": 0.035983758978545666,
219
  "reward_std": 0.11553369648754597,
 
229
  "kl": 0.169189453125,
230
  "learning_rate": 9.468163201617062e-05,
231
  "loss": -0.022672578692436218,
232
+ "memory(GiB)": 187.02,
233
  "step": 14,
234
  "train_speed(iter/s)": 0.000427
235
  },
 
241
  "kl": 0.166748046875,
242
  "learning_rate": 9.330127018922194e-05,
243
  "loss": -0.059799157083034515,
244
+ "memory(GiB)": 187.02,
245
  "response_clip_ratio": 0.4765625,
246
  "reward": 0.03584331553429365,
247
  "reward_std": 0.11829411797225475,
 
257
  "kl": 0.16748046875,
258
  "learning_rate": 9.177439057064683e-05,
259
  "loss": -0.06071458384394646,
260
+ "memory(GiB)": 187.02,
261
  "step": 16,
262
  "train_speed(iter/s)": 0.000431
263
  },
 
269
  "kl": 0.1787109375,
270
  "learning_rate": 9.01061596377522e-05,
271
  "loss": -0.04504441097378731,
272
+ "memory(GiB)": 187.02,
273
  "response_clip_ratio": 0.5625,
274
  "reward": 0.027318883687257767,
275
  "reward_std": 0.10441224090754986,
 
283
  "grad_norm": 0.005998397711664438,
284
  "learning_rate": 8.83022221559489e-05,
285
  "loss": -0.045487549155950546,
286
+ "memory(GiB)": 187.02,
287
  "step": 18,
288
  "train_speed(iter/s)": 0.000432
289
  },
 
311
  "kl": 0.1820068359375,
312
  "learning_rate": 8.636868207865244e-05,
313
  "loss": -0.03466903418302536,
314
+ "memory(GiB)": 187.02,
315
  "response_clip_ratio": 0.466796875,
316
  "reward": 0.04069916973821819,
317
  "reward_std": 0.11991005763411522,
 
327
  "kl": 0.19287109375,
328
  "learning_rate": 8.43120818934367e-05,
329
  "loss": -0.03502114117145538,
330
+ "memory(GiB)": 187.02,
331
  "step": 20,
332
  "train_speed(iter/s)": 0.000424
333
  },
 
339
  "kl": 0.17626953125,
340
  "learning_rate": 8.213938048432697e-05,
341
  "loss": -0.008662773296236992,
342
+ "memory(GiB)": 187.02,
343
  "response_clip_ratio": 0.5625,
344
  "reward": 0.04996980866417289,
345
  "reward_std": 0.13849420100450516,
 
355
  "kl": 0.178955078125,
356
  "learning_rate": 7.985792958513931e-05,
357
  "loss": -0.008743642829358578,
358
+ "memory(GiB)": 187.02,
359
  "step": 22,
360
  "train_speed(iter/s)": 0.000426
361
  },
 
367
  "kl": 0.1796875,
368
  "learning_rate": 7.74754489035403e-05,
369
  "loss": -0.03423420712351799,
370
+ "memory(GiB)": 187.02,
371
  "response_clip_ratio": 0.583984375,
372
  "reward": 0.034468831261619925,
373
  "reward_std": 0.11841745302081108,
 
381
  "grad_norm": 0.014131724834442139,
382
  "learning_rate": 7.500000000000001e-05,
383
  "loss": -0.03426633030176163,
384
+ "memory(GiB)": 187.02,
385
  "step": 24,
386
  "train_speed(iter/s)": 0.000427
387
  },
 
396
  "eval_reward_std": 0.10456253588199615,
397
  "eval_rewards/CosineReward": 0.04339282959699631,
398
  "eval_rewards/RepetitionPenalty": 0.0,
399
+ "eval_runtime": 1045.0642,
400
  "eval_samples_per_second": 0.001,
401
  "eval_steps_per_second": 0.001,
402
  "step": 24
 
409
  "kl": 0.1800537109375,
410
  "learning_rate": 7.243995901002312e-05,
411
  "loss": -0.02097315341234207,
412
+ "memory(GiB)": 187.02,
413
  "response_clip_ratio": 0.6171875,
414
  "reward": 0.03010205877944827,
415
  "reward_std": 0.10742511600255966,
 
425
  "kl": 0.18408203125,
426
  "learning_rate": 6.980398830195785e-05,
427
  "loss": -0.02103913575410843,
428
+ "memory(GiB)": 187.02,
429
  "step": 26,
430
  "train_speed(iter/s)": 0.000421
431
  },
 
437
  "kl": 0.174560546875,
438
  "learning_rate": 6.710100716628344e-05,
439
  "loss": -0.03593946248292923,
440
+ "memory(GiB)": 187.02,
441
  "response_clip_ratio": 0.513671875,
442
  "reward": 0.04752760287374258,
443
  "reward_std": 0.14935147762298584,
 
453
  "kl": 0.182373046875,
454
  "learning_rate": 6.434016163555452e-05,
455
  "loss": -0.03595500811934471,
456
+ "memory(GiB)": 187.02,
457
  "step": 28,
458
  "train_speed(iter/s)": 0.000422
459
  }
checkpoint-28/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7039ba231031019d68d36961179e879852b9f8a6de15562e5b792330bcbb4412
3
  size 9809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09cdf21dfd9faa218b7fd99e3f3dc0ef681c4e3fd3b905e7348f5467b0198044
3
  size 9809
checkpoint-30/README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: /mnt/nvme5n1p1/distill-14b-rl-70
3
+ library_name: peft
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
200
+ ### Framework versions
201
+
202
+ - PEFT 0.14.0
checkpoint-30/adapter_config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "/mnt/nvme5n1p1/distill-14b-rl-70",
5
+ "bias": "none",
6
+ "eva_config": null,
7
+ "exclude_modules": null,
8
+ "fan_in_fan_out": false,
9
+ "inference_mode": true,
10
+ "init_lora_weights": true,
11
+ "layer_replication": null,
12
+ "layers_pattern": null,
13
+ "layers_to_transform": null,
14
+ "loftq_config": {},
15
+ "lora_alpha": 32,
16
+ "lora_bias": false,
17
+ "lora_dropout": 0.05,
18
+ "megatron_config": null,
19
+ "megatron_core": "megatron.core",
20
+ "modules_to_save": [],
21
+ "peft_type": "LORA",
22
+ "r": 32,
23
+ "rank_pattern": {},
24
+ "revision": null,
25
+ "target_modules": [
26
+ "v_proj",
27
+ "up_proj",
28
+ "k_proj",
29
+ "gate_proj",
30
+ "o_proj",
31
+ "q_proj",
32
+ "down_proj"
33
+ ],
34
+ "task_type": "CAUSAL_LM",
35
+ "use_dora": false,
36
+ "use_rslora": false
37
+ }
checkpoint-30/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:329c17739679fed02726da82bfb68abbf7e99089c2e8a5c4d4e1f53724467219
3
+ size 275342392
checkpoint-30/additional_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}
checkpoint-30/global_step30/bf16_zero_pp_rank_16_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:236b7edd60ea120f67cd3fa17b6a0f26eafcb2b674178d748741a7bec780d3bb
3
+ size 51616527
checkpoint-30/global_step30/bf16_zero_pp_rank_17_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:577881ad19e84826a790e8334307e0604cacd87ff7f67e416aab193d1d3c7681
3
+ size 51616015
checkpoint-30/global_step30/bf16_zero_pp_rank_18_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e00a64034712680dbe3172fed4c3f9dac79b60a2559d4bae0ba510bbe197c59
3
+ size 51616527
checkpoint-30/global_step30/bf16_zero_pp_rank_19_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37b79763a682ec1a5717205a4c47851db99fb6ebeed1b00addf598d403c72fb2
3
+ size 51616015
checkpoint-30/global_step30/bf16_zero_pp_rank_20_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d637f1b940ee9ea28d165342c08527f8224e0f4d52fb8368a944a2120bd2a2d5
3
+ size 51616527
checkpoint-30/global_step30/bf16_zero_pp_rank_21_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d253294687babb38ecf1f3b1d6efb23f50026f15b6d92e2e6d55f500bb931dad
3
+ size 51616015
checkpoint-30/global_step30/bf16_zero_pp_rank_22_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:954697ae86ca3d7a1662c904d6d2e375b8348c71fc6fa59237b03f66fcf6f698
3
+ size 51616527
checkpoint-30/global_step30/bf16_zero_pp_rank_23_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:292c5cd764814ad0730375034dafd5b5316b951ea4b0d1a0c6d5f086e55cca1c
3
+ size 51616015
checkpoint-30/global_step30/mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19a396f183721ca31d21fa84e5530ebda200179fc3f57ea0b314e8f9828345e0
3
+ size 275768601
checkpoint-30/latest ADDED
@@ -0,0 +1 @@
 
 
1
+ global_step30
checkpoint-30/rng_state_16.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83cfe883ccee2993ebb068bb729f6106faeaa041208e83ceb6be530ff77d7075
3
+ size 16404
checkpoint-30/rng_state_17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:894b0396bee0214f53a66aaa33ee5095a9ef8c191fb926b5b8575a431b670362
3
+ size 16404
checkpoint-30/rng_state_18.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3126bd9635a55e5de0856de58320a3e0aca1161489085ae74ab449b1b8dd95b
3
+ size 16404