Back up Contextual Pixel 10% reproducibility probe
Browse files- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/optimizer.pt +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/pytorch_model.bin +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_0.pth +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_1.pth +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_2.pth +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_3.pth +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/scheduler.pt +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/trainer_state.json +616 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/training_args.bin +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/final/pytorch_model.bin +3 -0
- diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/final/training_complete.json +21 -0
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d08ef32845d4493b3dc99ce23a1c405c1a162f47f5945714c2f9a5f1d58e0aea
|
| 3 |
+
size 989585018
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39c89d9c796df1105be5b5d3a519c0bbcff2818e0775103a9aae3545d297333e
|
| 3 |
+
size 1942236130
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbe0d720c4c75a6a04213fa3b64bacbe794718a53e2b56ebb67a1a795014dfad
|
| 3 |
+
size 15024
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72452d3138d0ca2ff89429e3294a834ae7a68e8596fc757735ca56ae52509d57
|
| 3 |
+
size 15024
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f36e306fb8ebcf53a167bfd6c9af74db410a269ada1e619e3e816f5269543b9d
|
| 3 |
+
size 15024
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb47ce0c6f815a6f8302b0e3819b4c2315ca71dae3138d97fdceb765cdd0a039
|
| 3 |
+
size 15024
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae2307c96cdab87e8580b2205355cd09c43983cd01accb99baa0a8325e625964
|
| 3 |
+
size 1064
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.21294718909710392,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 6.382978723404255e-06,
|
| 13 |
+
"loss": 4.284,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.0,
|
| 18 |
+
"learning_rate": 1.276595744680851e-05,
|
| 19 |
+
"loss": 3.8831,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 1.9148936170212762e-05,
|
| 25 |
+
"loss": 3.6419,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 2.553191489361702e-05,
|
| 31 |
+
"loss": 3.5435,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 3.1914893617021275e-05,
|
| 37 |
+
"loss": 3.5036,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 3.8297872340425525e-05,
|
| 43 |
+
"loss": 3.4883,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01,
|
| 48 |
+
"learning_rate": 4.468085106382978e-05,
|
| 49 |
+
"loss": 3.4847,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.02,
|
| 54 |
+
"learning_rate": 5.106382978723404e-05,
|
| 55 |
+
"loss": 3.4292,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
+
"learning_rate": 5.7446808510638294e-05,
|
| 61 |
+
"loss": 3.4473,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
+
"learning_rate": 6.382978723404255e-05,
|
| 67 |
+
"loss": 3.443,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.02,
|
| 72 |
+
"learning_rate": 7.02127659574468e-05,
|
| 73 |
+
"loss": 3.4067,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.03,
|
| 78 |
+
"learning_rate": 7.659574468085105e-05,
|
| 79 |
+
"loss": 3.3916,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.03,
|
| 84 |
+
"learning_rate": 8.297872340425531e-05,
|
| 85 |
+
"loss": 3.4239,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03,
|
| 90 |
+
"learning_rate": 8.936170212765956e-05,
|
| 91 |
+
"loss": 3.4058,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.03,
|
| 96 |
+
"learning_rate": 9.574468085106382e-05,
|
| 97 |
+
"loss": 3.3709,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
+
"learning_rate": 0.00010212765957446807,
|
| 103 |
+
"loss": 3.3949,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.04,
|
| 108 |
+
"learning_rate": 0.00010851063829787234,
|
| 109 |
+
"loss": 3.4156,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.04,
|
| 114 |
+
"learning_rate": 0.00011489361702127659,
|
| 115 |
+
"loss": 3.3717,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.04,
|
| 120 |
+
"learning_rate": 0.00012127659574468084,
|
| 121 |
+
"loss": 3.3702,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04,
|
| 126 |
+
"learning_rate": 0.0001276595744680851,
|
| 127 |
+
"loss": 3.3759,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04,
|
| 132 |
+
"learning_rate": 0.00013404255319148935,
|
| 133 |
+
"loss": 3.3641,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.05,
|
| 138 |
+
"learning_rate": 0.0001404255319148936,
|
| 139 |
+
"loss": 3.3645,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.05,
|
| 144 |
+
"learning_rate": 0.00014680851063829785,
|
| 145 |
+
"loss": 3.3668,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.05,
|
| 150 |
+
"learning_rate": 0.0001531914893617021,
|
| 151 |
+
"loss": 3.3511,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.05,
|
| 156 |
+
"learning_rate": 0.00015957446808510637,
|
| 157 |
+
"loss": 3.3755,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.06,
|
| 162 |
+
"learning_rate": 0.00016595744680851062,
|
| 163 |
+
"loss": 3.3214,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.06,
|
| 168 |
+
"learning_rate": 0.0001723404255319149,
|
| 169 |
+
"loss": 3.3533,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.06,
|
| 174 |
+
"learning_rate": 0.00017872340425531912,
|
| 175 |
+
"loss": 3.3422,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.06,
|
| 180 |
+
"learning_rate": 0.0001851063829787234,
|
| 181 |
+
"loss": 3.3571,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.06,
|
| 186 |
+
"learning_rate": 0.00019148936170212765,
|
| 187 |
+
"loss": 3.3484,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.07,
|
| 192 |
+
"learning_rate": 0.00019787234042553187,
|
| 193 |
+
"loss": 3.3172,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.07,
|
| 198 |
+
"learning_rate": 0.00020425531914893615,
|
| 199 |
+
"loss": 3.3343,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.07,
|
| 204 |
+
"learning_rate": 0.0002106382978723404,
|
| 205 |
+
"loss": 3.3229,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.07,
|
| 210 |
+
"learning_rate": 0.00021702127659574468,
|
| 211 |
+
"loss": 3.3166,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.07,
|
| 216 |
+
"learning_rate": 0.0002234042553191489,
|
| 217 |
+
"loss": 3.3227,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.08,
|
| 222 |
+
"learning_rate": 0.00022978723404255317,
|
| 223 |
+
"loss": 3.3401,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.08,
|
| 228 |
+
"learning_rate": 0.00023617021276595742,
|
| 229 |
+
"loss": 3.3182,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.08,
|
| 234 |
+
"learning_rate": 0.00024255319148936167,
|
| 235 |
+
"loss": 3.3517,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.08,
|
| 240 |
+
"learning_rate": 0.0002489361702127659,
|
| 241 |
+
"loss": 3.3366,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.09,
|
| 246 |
+
"learning_rate": 0.0002553191489361702,
|
| 247 |
+
"loss": 3.2966,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.09,
|
| 252 |
+
"learning_rate": 0.0002617021276595745,
|
| 253 |
+
"loss": 3.3101,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.09,
|
| 258 |
+
"learning_rate": 0.0002680851063829787,
|
| 259 |
+
"loss": 3.2936,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.09,
|
| 264 |
+
"learning_rate": 0.000274468085106383,
|
| 265 |
+
"loss": 3.3299,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.09,
|
| 270 |
+
"learning_rate": 0.0002808510638297872,
|
| 271 |
+
"loss": 3.3214,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.1,
|
| 276 |
+
"learning_rate": 0.0002872340425531915,
|
| 277 |
+
"loss": 3.3319,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.1,
|
| 282 |
+
"learning_rate": 0.0002936170212765957,
|
| 283 |
+
"loss": 3.3418,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.1,
|
| 288 |
+
"learning_rate": 0.0003,
|
| 289 |
+
"loss": 3.3488,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.1,
|
| 294 |
+
"learning_rate": 0.0002999958552339374,
|
| 295 |
+
"loss": 3.3235,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.1,
|
| 300 |
+
"learning_rate": 0.0002999834211648041,
|
| 301 |
+
"loss": 3.2771,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.11,
|
| 306 |
+
"learning_rate": 0.00029996269847975095,
|
| 307 |
+
"loss": 3.29,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.11,
|
| 312 |
+
"learning_rate": 0.0002999336883239869,
|
| 313 |
+
"loss": 3.3109,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.11,
|
| 318 |
+
"learning_rate": 0.0002998963923007162,
|
| 319 |
+
"loss": 3.2893,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.11,
|
| 324 |
+
"learning_rate": 0.0002998508124710493,
|
| 325 |
+
"loss": 3.2992,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.11,
|
| 330 |
+
"learning_rate": 0.00029979695135388934,
|
| 331 |
+
"loss": 3.2709,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.12,
|
| 336 |
+
"learning_rate": 0.0002997348119257927,
|
| 337 |
+
"loss": 3.2941,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.12,
|
| 342 |
+
"learning_rate": 0.00029966439762080466,
|
| 343 |
+
"loss": 3.2795,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.12,
|
| 348 |
+
"learning_rate": 0.0002995857123302695,
|
| 349 |
+
"loss": 3.279,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.12,
|
| 354 |
+
"learning_rate": 0.0002994987604026154,
|
| 355 |
+
"loss": 3.2673,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.13,
|
| 360 |
+
"learning_rate": 0.0002994035466431145,
|
| 361 |
+
"loss": 3.291,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.13,
|
| 366 |
+
"learning_rate": 0.0002993000763136169,
|
| 367 |
+
"loss": 3.2783,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.13,
|
| 372 |
+
"learning_rate": 0.00029918835513226,
|
| 373 |
+
"loss": 3.2758,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.13,
|
| 378 |
+
"learning_rate": 0.0002990683892731526,
|
| 379 |
+
"loss": 3.2683,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.13,
|
| 384 |
+
"learning_rate": 0.0002989401853660337,
|
| 385 |
+
"loss": 3.3042,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.14,
|
| 390 |
+
"learning_rate": 0.00029880375049590597,
|
| 391 |
+
"loss": 3.2644,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.14,
|
| 396 |
+
"learning_rate": 0.0002986590922026444,
|
| 397 |
+
"loss": 3.2687,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.14,
|
| 402 |
+
"learning_rate": 0.0002985062184805794,
|
| 403 |
+
"loss": 3.2694,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.14,
|
| 408 |
+
"learning_rate": 0.00029834513777805523,
|
| 409 |
+
"loss": 3.2757,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.14,
|
| 414 |
+
"learning_rate": 0.0002981758589969629,
|
| 415 |
+
"loss": 3.269,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.15,
|
| 420 |
+
"learning_rate": 0.00029799839149224834,
|
| 421 |
+
"loss": 3.2642,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.15,
|
| 426 |
+
"learning_rate": 0.0002978127450713955,
|
| 427 |
+
"loss": 3.2502,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.15,
|
| 432 |
+
"learning_rate": 0.0002976189299938841,
|
| 433 |
+
"loss": 3.2641,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.15,
|
| 438 |
+
"learning_rate": 0.0002974169569706229,
|
| 439 |
+
"loss": 3.252,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.16,
|
| 444 |
+
"learning_rate": 0.00029720683716335773,
|
| 445 |
+
"loss": 3.2828,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.16,
|
| 450 |
+
"learning_rate": 0.00029698858218405453,
|
| 451 |
+
"loss": 3.2763,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.16,
|
| 456 |
+
"learning_rate": 0.00029676220409425766,
|
| 457 |
+
"loss": 3.2816,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.16,
|
| 462 |
+
"learning_rate": 0.00029652771540442346,
|
| 463 |
+
"loss": 3.2652,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.16,
|
| 468 |
+
"learning_rate": 0.00029628512907322875,
|
| 469 |
+
"loss": 3.2723,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.17,
|
| 474 |
+
"learning_rate": 0.0002960344585068548,
|
| 475 |
+
"loss": 3.2669,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.17,
|
| 480 |
+
"learning_rate": 0.00029577571755824635,
|
| 481 |
+
"loss": 3.2657,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.17,
|
| 486 |
+
"learning_rate": 0.0002955089205263461,
|
| 487 |
+
"loss": 3.2691,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.17,
|
| 492 |
+
"learning_rate": 0.00029523408215530454,
|
| 493 |
+
"loss": 3.2589,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.17,
|
| 498 |
+
"learning_rate": 0.00029495121763366505,
|
| 499 |
+
"loss": 3.2575,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.18,
|
| 504 |
+
"learning_rate": 0.0002946603425935245,
|
| 505 |
+
"loss": 3.2689,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.18,
|
| 510 |
+
"learning_rate": 0.0002943614731096694,
|
| 511 |
+
"loss": 3.2592,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.18,
|
| 516 |
+
"learning_rate": 0.00029405462569868793,
|
| 517 |
+
"loss": 3.2739,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.18,
|
| 522 |
+
"learning_rate": 0.00029373981731805633,
|
| 523 |
+
"loss": 3.2887,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.19,
|
| 528 |
+
"learning_rate": 0.0002934170653652026,
|
| 529 |
+
"loss": 3.2776,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.19,
|
| 534 |
+
"learning_rate": 0.0002930863876765446,
|
| 535 |
+
"loss": 3.2569,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.19,
|
| 540 |
+
"learning_rate": 0.0002927478025265044,
|
| 541 |
+
"loss": 3.2435,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.19,
|
| 546 |
+
"learning_rate": 0.00029240132862649865,
|
| 547 |
+
"loss": 3.2493,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.19,
|
| 552 |
+
"learning_rate": 0.0002920469851239041,
|
| 553 |
+
"loss": 3.2596,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.2,
|
| 558 |
+
"learning_rate": 0.00029168479160099987,
|
| 559 |
+
"loss": 3.2698,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.2,
|
| 564 |
+
"learning_rate": 0.0002913147680738847,
|
| 565 |
+
"loss": 3.2378,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.2,
|
| 570 |
+
"learning_rate": 0.00029093693499137154,
|
| 571 |
+
"loss": 3.2727,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.2,
|
| 576 |
+
"learning_rate": 0.0002905513132338568,
|
| 577 |
+
"loss": 3.2405,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.2,
|
| 582 |
+
"learning_rate": 0.0002901579241121668,
|
| 583 |
+
"loss": 3.2417,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.21,
|
| 588 |
+
"learning_rate": 0.00028975678936638,
|
| 589 |
+
"loss": 3.2505,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.21,
|
| 594 |
+
"learning_rate": 0.00028934793116462546,
|
| 595 |
+
"loss": 3.2443,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.21,
|
| 600 |
+
"learning_rate": 0.0002889313721018578,
|
| 601 |
+
"loss": 3.2423,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.21,
|
| 606 |
+
"learning_rate": 0.0002885071351986086,
|
| 607 |
+
"loss": 3.2409,
|
| 608 |
+
"step": 1000
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 4696,
|
| 612 |
+
"num_train_epochs": 1,
|
| 613 |
+
"total_flos": 0.0,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d4f45c5c8adeb3c88fe8a0e8918e592c26472924156695546e77bde30f0fea50
|
| 3 |
+
size 4536
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/final/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3ac75fd4c8132831191b5e203a2450a18892756b6358a88f9eaa0736149be26
|
| 3 |
+
size 1942239884
|
diagnostics/repro_contextual_pct10/checkpoints/alignment_probe_1000/final/training_complete.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epochs": 1,
|
| 3 |
+
"global_step": 1000,
|
| 4 |
+
"initial_weights": null,
|
| 5 |
+
"learning_rate": 0.0003,
|
| 6 |
+
"metrics": {
|
| 7 |
+
"epoch": 0.21,
|
| 8 |
+
"train_loss": 3.334453639984131,
|
| 9 |
+
"train_runtime": 553.9729,
|
| 10 |
+
"train_samples_per_second": 4339.931,
|
| 11 |
+
"train_steps_per_second": 8.477
|
| 12 |
+
},
|
| 13 |
+
"model_type": "pixel",
|
| 14 |
+
"percentage": 10,
|
| 15 |
+
"scheduler": "cosine",
|
| 16 |
+
"seed": 42,
|
| 17 |
+
"stage": "alignment",
|
| 18 |
+
"subset_rows": 2404204,
|
| 19 |
+
"total_rows": 24042041,
|
| 20 |
+
"warmup_ratio": 0.1
|
| 21 |
+
}
|