Upload folder using huggingface_hub
Browse files
RuleMaze/quest/checkpoint/adapter_config.json
CHANGED
|
@@ -29,118 +29,118 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
-
"layers.
|
| 33 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
| 34 |
"layers.16.mlp.down_proj",
|
| 35 |
-
"layers.12.mlp.down_proj",
|
| 36 |
-
"layers.31.mlp.up_proj",
|
| 37 |
-
"layers.12.mlp.up_proj",
|
| 38 |
-
"layers.20.mlp.gate_proj",
|
| 39 |
-
"layers.4.mlp.up_proj",
|
| 40 |
-
"layers.11.mlp.up_proj",
|
| 41 |
-
"layers.0.mlp.down_proj",
|
| 42 |
-
"layers.17.mlp.down_proj",
|
| 43 |
-
"layers.29.mlp.down_proj",
|
| 44 |
-
"layers.19.mlp.down_proj",
|
| 45 |
-
"layers.4.mlp.gate_proj",
|
| 46 |
-
"layers.26.mlp.up_proj",
|
| 47 |
-
"layers.22.mlp.up_proj",
|
| 48 |
-
"layers.0.mlp.gate_proj",
|
| 49 |
-
"layers.24.mlp.down_proj",
|
| 50 |
-
"layers.23.mlp.down_proj",
|
| 51 |
-
"layers.25.mlp.up_proj",
|
| 52 |
-
"35.mlp.gate_proj",
|
| 53 |
-
"layers.21.mlp.gate_proj",
|
| 54 |
"layers.6.mlp.up_proj",
|
|
|
|
| 55 |
"layers.10.mlp.gate_proj",
|
| 56 |
-
"layers.18.mlp.up_proj",
|
| 57 |
"35.mlp.up_proj",
|
| 58 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
"layers.13.mlp.gate_proj",
|
| 60 |
-
"
|
| 61 |
-
"layers.
|
| 62 |
-
"layers.
|
| 63 |
-
"layers.
|
| 64 |
-
"layers.
|
| 65 |
-
"layers.
|
| 66 |
-
"layers.19.mlp.up_proj",
|
| 67 |
-
"layers.2.mlp.up_proj",
|
| 68 |
"layers.9.mlp.gate_proj",
|
| 69 |
-
"
|
| 70 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
"layers.6.mlp.down_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
"o_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 73 |
"layers.3.mlp.up_proj",
|
| 74 |
-
"layers.
|
| 75 |
"layers.5.mlp.gate_proj",
|
| 76 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 77 |
"layers.10.mlp.down_proj",
|
| 78 |
-
"layers.27.mlp.gate_proj",
|
| 79 |
"33.mlp.gate_proj",
|
| 80 |
-
"layers.
|
| 81 |
-
"layers.
|
| 82 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 83 |
"layers.1.mlp.up_proj",
|
| 84 |
-
"
|
| 85 |
-
"
|
| 86 |
-
"layers.
|
|
|
|
| 87 |
"layers.26.mlp.down_proj",
|
| 88 |
-
"
|
| 89 |
-
"layers.
|
| 90 |
-
"layers.
|
| 91 |
-
"
|
| 92 |
-
"layers.3.mlp.down_proj",
|
| 93 |
-
"layers.8.mlp.down_proj",
|
| 94 |
-
"layers.16.mlp.gate_proj",
|
| 95 |
"35.mlp.down_proj",
|
| 96 |
-
"
|
| 97 |
-
"layers.
|
| 98 |
-
"layers.19.mlp.gate_proj",
|
| 99 |
-
"layers.14.mlp.down_proj",
|
| 100 |
"33.mlp.down_proj",
|
| 101 |
-
"
|
|
|
|
| 102 |
"layers.18.mlp.down_proj",
|
| 103 |
-
"layers.
|
| 104 |
-
"layers.27.mlp.down_proj",
|
| 105 |
"layers.17.mlp.up_proj",
|
| 106 |
-
"layers.
|
| 107 |
-
"layers.27.mlp.up_proj",
|
| 108 |
-
"layers.1.mlp.down_proj",
|
| 109 |
-
"layers.12.mlp.gate_proj",
|
| 110 |
-
"layers.1.mlp.gate_proj",
|
| 111 |
-
"layers.8.mlp.gate_proj",
|
| 112 |
-
"layers.15.mlp.up_proj",
|
| 113 |
-
"layers.3.mlp.gate_proj",
|
| 114 |
-
"layers.24.mlp.gate_proj",
|
| 115 |
-
"layers.29.mlp.up_proj",
|
| 116 |
-
"layers.11.mlp.down_proj",
|
| 117 |
-
"layers.23.mlp.gate_proj",
|
| 118 |
-
"layers.7.mlp.down_proj",
|
| 119 |
-
"layers.13.mlp.up_proj",
|
| 120 |
-
"layers.7.mlp.up_proj",
|
| 121 |
-
"layers.26.mlp.gate_proj",
|
| 122 |
-
"layers.17.mlp.gate_proj",
|
| 123 |
-
"layers.24.mlp.up_proj",
|
| 124 |
-
"layers.23.mlp.up_proj",
|
| 125 |
-
"layers.5.mlp.down_proj",
|
| 126 |
-
"layers.6.mlp.gate_proj",
|
| 127 |
-
"layers.5.mlp.up_proj",
|
| 128 |
-
"layers.2.mlp.down_proj",
|
| 129 |
-
"layers.15.mlp.gate_proj",
|
| 130 |
-
"layers.9.mlp.up_proj",
|
| 131 |
-
"layers.2.mlp.gate_proj",
|
| 132 |
-
"32.mlp.down_proj",
|
| 133 |
-
"layers.25.mlp.gate_proj",
|
| 134 |
-
"32.mlp.gate_proj",
|
| 135 |
-
"layers.18.mlp.gate_proj",
|
| 136 |
-
"layers.28.mlp.down_proj",
|
| 137 |
-
"layers.0.mlp.up_proj",
|
| 138 |
-
"layers.16.mlp.up_proj",
|
| 139 |
-
"layers.15.mlp.down_proj",
|
| 140 |
-
"layers.29.mlp.gate_proj",
|
| 141 |
-
"34.mlp.gate_proj",
|
| 142 |
-
"k_proj",
|
| 143 |
-
"layers.11.mlp.gate_proj"
|
| 144 |
],
|
| 145 |
"target_parameters": null,
|
| 146 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"layers.13.mlp.up_proj",
|
| 33 |
+
"layers.20.mlp.up_proj",
|
| 34 |
+
"layers.23.mlp.up_proj",
|
| 35 |
+
"layers.8.mlp.gate_proj",
|
| 36 |
+
"layers.0.mlp.up_proj",
|
| 37 |
"layers.16.mlp.down_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
"layers.6.mlp.up_proj",
|
| 39 |
+
"layers.19.mlp.gate_proj",
|
| 40 |
"layers.10.mlp.gate_proj",
|
|
|
|
| 41 |
"35.mlp.up_proj",
|
| 42 |
+
"layers.3.mlp.down_proj",
|
| 43 |
+
"layers.29.mlp.down_proj",
|
| 44 |
+
"layers.23.mlp.down_proj",
|
| 45 |
+
"layers.6.mlp.gate_proj",
|
| 46 |
+
"layers.2.mlp.down_proj",
|
| 47 |
"layers.13.mlp.gate_proj",
|
| 48 |
+
"k_proj",
|
| 49 |
+
"layers.1.mlp.down_proj",
|
| 50 |
+
"layers.4.mlp.up_proj",
|
| 51 |
+
"layers.21.mlp.up_proj",
|
| 52 |
+
"layers.28.mlp.gate_proj",
|
| 53 |
+
"layers.12.mlp.down_proj",
|
|
|
|
|
|
|
| 54 |
"layers.9.mlp.gate_proj",
|
| 55 |
+
"layers.29.mlp.up_proj",
|
| 56 |
+
"34.mlp.down_proj",
|
| 57 |
+
"layers.15.mlp.gate_proj",
|
| 58 |
+
"layers.9.mlp.up_proj",
|
| 59 |
+
"layers.21.mlp.gate_proj",
|
| 60 |
+
"34.mlp.up_proj",
|
| 61 |
+
"layers.7.mlp.gate_proj",
|
| 62 |
+
"layers.27.mlp.down_proj",
|
| 63 |
"layers.6.mlp.down_proj",
|
| 64 |
+
"32.mlp.gate_proj",
|
| 65 |
+
"layers.8.mlp.up_proj",
|
| 66 |
+
"layers.12.mlp.gate_proj",
|
| 67 |
+
"layers.25.mlp.up_proj",
|
| 68 |
+
"layers.31.mlp.up_proj",
|
| 69 |
+
"layers.14.mlp.gate_proj",
|
| 70 |
+
"layers.2.mlp.up_proj",
|
| 71 |
+
"layers.5.mlp.down_proj",
|
| 72 |
+
"layers.28.mlp.down_proj",
|
| 73 |
+
"layers.22.mlp.up_proj",
|
| 74 |
+
"layers.11.mlp.down_proj",
|
| 75 |
+
"layers.8.mlp.down_proj",
|
| 76 |
+
"layers.0.mlp.gate_proj",
|
| 77 |
+
"layers.25.mlp.down_proj",
|
| 78 |
"o_proj",
|
| 79 |
+
"layers.14.mlp.up_proj",
|
| 80 |
+
"layers.19.mlp.down_proj",
|
| 81 |
+
"layers.20.mlp.gate_proj",
|
| 82 |
+
"layers.15.mlp.down_proj",
|
| 83 |
+
"layers.22.mlp.down_proj",
|
| 84 |
+
"layers.16.mlp.up_proj",
|
| 85 |
+
"layers.30.mlp.down_proj",
|
| 86 |
+
"layers.22.mlp.gate_proj",
|
| 87 |
+
"34.mlp.gate_proj",
|
| 88 |
+
"layers.0.mlp.down_proj",
|
| 89 |
+
"layers.21.mlp.down_proj",
|
| 90 |
+
"32.mlp.down_proj",
|
| 91 |
+
"layers.26.mlp.gate_proj",
|
| 92 |
+
"layers.25.mlp.gate_proj",
|
| 93 |
+
"layers.11.mlp.up_proj",
|
| 94 |
+
"q_proj",
|
| 95 |
+
"layers.17.mlp.down_proj",
|
| 96 |
+
"layers.19.mlp.up_proj",
|
| 97 |
+
"layers.2.mlp.gate_proj",
|
| 98 |
+
"layers.24.mlp.up_proj",
|
| 99 |
+
"layers.31.mlp.gate_proj",
|
| 100 |
+
"layers.4.mlp.gate_proj",
|
| 101 |
+
"layers.13.mlp.down_proj",
|
| 102 |
"layers.3.mlp.up_proj",
|
| 103 |
+
"layers.5.mlp.up_proj",
|
| 104 |
"layers.5.mlp.gate_proj",
|
| 105 |
+
"layers.11.mlp.gate_proj",
|
| 106 |
+
"layers.30.mlp.gate_proj",
|
| 107 |
+
"32.mlp.up_proj",
|
| 108 |
+
"layers.9.mlp.down_proj",
|
| 109 |
+
"layers.28.mlp.up_proj",
|
| 110 |
+
"33.mlp.up_proj",
|
| 111 |
+
"layers.24.mlp.down_proj",
|
| 112 |
+
"layers.24.mlp.gate_proj",
|
| 113 |
+
"layers.16.mlp.gate_proj",
|
| 114 |
+
"layers.12.mlp.up_proj",
|
| 115 |
"layers.10.mlp.down_proj",
|
|
|
|
| 116 |
"33.mlp.gate_proj",
|
| 117 |
+
"layers.27.mlp.gate_proj",
|
| 118 |
+
"layers.31.mlp.down_proj",
|
| 119 |
+
"35.mlp.gate_proj",
|
| 120 |
+
"layers.15.mlp.up_proj",
|
| 121 |
+
"layers.10.mlp.up_proj",
|
| 122 |
+
"layers.7.mlp.up_proj",
|
| 123 |
+
"layers.3.mlp.gate_proj",
|
| 124 |
"layers.1.mlp.up_proj",
|
| 125 |
+
"layers.1.mlp.gate_proj",
|
| 126 |
+
"layers.27.mlp.up_proj",
|
| 127 |
+
"layers.29.mlp.gate_proj",
|
| 128 |
+
"layers.14.mlp.down_proj",
|
| 129 |
"layers.26.mlp.down_proj",
|
| 130 |
+
"layers.18.mlp.up_proj",
|
| 131 |
+
"layers.4.mlp.down_proj",
|
| 132 |
+
"layers.17.mlp.gate_proj",
|
| 133 |
+
"layers.23.mlp.gate_proj",
|
|
|
|
|
|
|
|
|
|
| 134 |
"35.mlp.down_proj",
|
| 135 |
+
"layers.26.mlp.up_proj",
|
| 136 |
+
"layers.18.mlp.gate_proj",
|
|
|
|
|
|
|
| 137 |
"33.mlp.down_proj",
|
| 138 |
+
"v_proj",
|
| 139 |
+
"layers.7.mlp.down_proj",
|
| 140 |
"layers.18.mlp.down_proj",
|
| 141 |
+
"layers.30.mlp.up_proj",
|
|
|
|
| 142 |
"layers.17.mlp.up_proj",
|
| 143 |
+
"layers.20.mlp.down_proj"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 144 |
],
|
| 145 |
"target_parameters": null,
|
| 146 |
"task_type": "CAUSAL_LM",
|
RuleMaze/quest/checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 59941192
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ec21d4212d54e6e906db3a43e3fbdb092c7d31bae36687d6519ca8d9bddebc9
|
| 3 |
size 59941192
|
RuleMaze/quest/checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 120158586
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19de3bfadfa40d10b2eabffa786b0e884e4b2ef2457f0f1c3197971cf8cd8ab0
|
| 3 |
size 120158586
|
RuleMaze/quest/checkpoint/trainer_state.json
CHANGED
|
@@ -11,874 +11,874 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.06901311249137336,
|
| 14 |
-
"grad_norm": 1.
|
| 15 |
"learning_rate": 3.4161490683229816e-06,
|
| 16 |
-
"loss": 0.
|
| 17 |
"step": 100
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.13802622498274672,
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
"learning_rate": 6.86680469289165e-06,
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 200
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2070393374741201,
|
| 28 |
-
"grad_norm": 1.
|
| 29 |
"learning_rate": 1.0317460317460318e-05,
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 300
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.27605244996549344,
|
| 35 |
-
"grad_norm": 1.
|
| 36 |
"learning_rate": 1.3768115942028985e-05,
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 400
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.3450655624568668,
|
| 42 |
-
"grad_norm": 1.
|
| 43 |
"learning_rate": 1.7218771566597654e-05,
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.4140786749482402,
|
| 49 |
-
"grad_norm":
|
| 50 |
"learning_rate": 2.0669427191166323e-05,
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 600
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.4830917874396135,
|
| 56 |
-
"grad_norm":
|
| 57 |
"learning_rate": 2.412008281573499e-05,
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 700
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.5521048999309869,
|
| 63 |
-
"grad_norm": 0.
|
| 64 |
"learning_rate": 2.757073844030366e-05,
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 800
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.6211180124223602,
|
| 70 |
-
"grad_norm": 1.
|
| 71 |
"learning_rate": 3.102139406487233e-05,
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 900
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.6901311249137336,
|
| 77 |
-
"grad_norm":
|
| 78 |
"learning_rate": 3.4472049689440996e-05,
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 1000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.759144237405107,
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
"learning_rate": 3.7922705314009665e-05,
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 1100
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.8281573498964804,
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
"learning_rate": 4.1373360938578333e-05,
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 1200
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.8971704623878537,
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
"learning_rate": 4.4824016563146996e-05,
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 1300
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.966183574879227,
|
| 105 |
-
"grad_norm":
|
| 106 |
"learning_rate": 4.827467218771567e-05,
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 1400
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 1.0351966873706004,
|
| 112 |
-
"grad_norm": 1.
|
| 113 |
"learning_rate": 5.172532781228434e-05,
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 1500
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 1.1042097998619738,
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
"learning_rate": 5.517598343685301e-05,
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 1600
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 1.173222912353347,
|
| 126 |
-
"grad_norm": 1.
|
| 127 |
"learning_rate": 5.862663906142167e-05,
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 1700
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 1.2422360248447206,
|
| 133 |
-
"grad_norm": 1.
|
| 134 |
"learning_rate": 6.207729468599034e-05,
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 1800
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 1.311249137336094,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
"learning_rate": 6.5527950310559e-05,
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 1900
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 1.3802622498274673,
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
"learning_rate": 6.897860593512768e-05,
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 2000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 1.4492753623188406,
|
| 154 |
-
"grad_norm":
|
| 155 |
"learning_rate": 7.242926155969635e-05,
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 2100
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 1.518288474810214,
|
| 161 |
-
"grad_norm":
|
| 162 |
"learning_rate": 7.587991718426502e-05,
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 2200
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 1.5873015873015874,
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
"learning_rate": 7.933057280883368e-05,
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 2300
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 1.6563146997929605,
|
| 175 |
-
"grad_norm": 1.
|
| 176 |
"learning_rate": 8.278122843340234e-05,
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 2400
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 1.725327812284334,
|
| 182 |
-
"grad_norm": 2.
|
| 183 |
"learning_rate": 8.623188405797103e-05,
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 2500
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 1.7943409247757074,
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
"learning_rate": 8.968253968253969e-05,
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 2600
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 1.8633540372670807,
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
"learning_rate": 9.313319530710835e-05,
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 2700
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 1.9323671497584543,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
"learning_rate": 9.658385093167702e-05,
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 2800
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 2.0013802622498273,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
"learning_rate": 9.99999996372913e-05,
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 2900
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 2.070393374741201,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
"learning_rate": 9.999630005418305e-05,
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 3000
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 2.070393374741201,
|
| 224 |
-
"eval_loss": 0.
|
| 225 |
-
"eval_runtime":
|
| 226 |
-
"eval_samples_per_second":
|
| 227 |
-
"eval_steps_per_second":
|
| 228 |
"step": 3000
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 2.139406487232574,
|
| 232 |
-
"grad_norm": 0.
|
| 233 |
"learning_rate": 9.998534692157168e-05,
|
| 234 |
-
"loss": 0.
|
| 235 |
"step": 3100
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 2.2084195997239475,
|
| 239 |
-
"grad_norm": 0.
|
| 240 |
"learning_rate": 9.99671418285566e-05,
|
| 241 |
-
"loss": 0.
|
| 242 |
"step": 3200
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 2.277432712215321,
|
| 246 |
-
"grad_norm": 0.
|
| 247 |
"learning_rate": 9.994168741636414e-05,
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 3300
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 2.346445824706694,
|
| 253 |
-
"grad_norm": 0.
|
| 254 |
"learning_rate": 9.990898737796431e-05,
|
| 255 |
-
"loss": 0.
|
| 256 |
"step": 3400
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 2.4154589371980677,
|
| 260 |
-
"grad_norm": 0.
|
| 261 |
"learning_rate": 9.986904645753517e-05,
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 3500
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 2.4844720496894412,
|
| 267 |
-
"grad_norm": 0.
|
| 268 |
"learning_rate": 9.982187044977433e-05,
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 3600
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 2.5534851621808143,
|
| 274 |
-
"grad_norm": 0.
|
| 275 |
"learning_rate": 9.976746619905848e-05,
|
| 276 |
-
"loss": 0.
|
| 277 |
"step": 3700
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 2.622498274672188,
|
| 281 |
-
"grad_norm": 0.
|
| 282 |
"learning_rate": 9.970584159845019e-05,
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 3800
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 2.691511387163561,
|
| 288 |
-
"grad_norm": 0.
|
| 289 |
"learning_rate": 9.963700558855286e-05,
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 3900
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 2.7605244996549345,
|
| 295 |
-
"grad_norm": 0.
|
| 296 |
"learning_rate": 9.956096815621365e-05,
|
| 297 |
-
"loss": 0.
|
| 298 |
"step": 4000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 2.829537612146308,
|
| 302 |
-
"grad_norm": 0.
|
| 303 |
"learning_rate": 9.947774033307444e-05,
|
| 304 |
-
"loss": 0.
|
| 305 |
"step": 4100
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 2.898550724637681,
|
| 309 |
-
"grad_norm": 0.
|
| 310 |
"learning_rate": 9.938733419397148e-05,
|
| 311 |
-
"loss": 0.
|
| 312 |
"step": 4200
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 2.9675638371290542,
|
| 316 |
-
"grad_norm": 0.
|
| 317 |
"learning_rate": 9.928976285518347e-05,
|
| 318 |
-
"loss": 0.
|
| 319 |
"step": 4300
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 3.036576949620428,
|
| 323 |
-
"grad_norm": 0.
|
| 324 |
"learning_rate": 9.918504047252866e-05,
|
| 325 |
-
"loss": 0.
|
| 326 |
"step": 4400
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 3.1055900621118013,
|
| 330 |
-
"grad_norm": 0.
|
| 331 |
"learning_rate": 9.907318223931104e-05,
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 4500
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 3.1746031746031744,
|
| 337 |
-
"grad_norm": 0.
|
| 338 |
"learning_rate": 9.895420438411616e-05,
|
| 339 |
-
"loss": 0.
|
| 340 |
"step": 4600
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 3.243616287094548,
|
| 344 |
-
"grad_norm": 0.
|
| 345 |
"learning_rate": 9.882812416845659e-05,
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 4700
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 3.3126293995859215,
|
| 351 |
-
"grad_norm": 0.
|
| 352 |
"learning_rate": 9.869495988426765e-05,
|
| 353 |
-
"loss": 0.
|
| 354 |
"step": 4800
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 3.3816425120772946,
|
| 358 |
-
"grad_norm": 0.
|
| 359 |
"learning_rate": 9.85547308512535e-05,
|
| 360 |
-
"loss": 0.
|
| 361 |
"step": 4900
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 3.450655624568668,
|
| 365 |
-
"grad_norm": 0.
|
| 366 |
"learning_rate": 9.84074574140843e-05,
|
| 367 |
-
"loss": 0.
|
| 368 |
"step": 5000
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 3.5196687370600412,
|
| 372 |
-
"grad_norm": 0.
|
| 373 |
"learning_rate": 9.825316093944449e-05,
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 5100
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 3.588681849551415,
|
| 379 |
-
"grad_norm": 0.
|
| 380 |
"learning_rate": 9.809186381293293e-05,
|
| 381 |
-
"loss": 0.
|
| 382 |
"step": 5200
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 3.657694962042788,
|
| 386 |
-
"grad_norm": 0.
|
| 387 |
"learning_rate": 9.792358943581512e-05,
|
| 388 |
-
"loss": 0.
|
| 389 |
"step": 5300
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 3.7267080745341614,
|
| 393 |
-
"grad_norm": 0.
|
| 394 |
"learning_rate": 9.774836222162811e-05,
|
| 395 |
-
"loss": 0.
|
| 396 |
"step": 5400
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 3.795721187025535,
|
| 400 |
-
"grad_norm": 0.
|
| 401 |
"learning_rate": 9.756620759263859e-05,
|
| 402 |
-
"loss": 0.
|
| 403 |
"step": 5500
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 3.864734299516908,
|
| 407 |
-
"grad_norm": 0.
|
| 408 |
"learning_rate": 9.73771519761545e-05,
|
| 409 |
-
"loss": 0.
|
| 410 |
"step": 5600
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 3.9337474120082816,
|
| 414 |
-
"grad_norm": 0.
|
| 415 |
"learning_rate": 9.718122280069101e-05,
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 5700
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 4.002760524499655,
|
| 421 |
-
"grad_norm": 0.
|
| 422 |
"learning_rate": 9.697844849199104e-05,
|
| 423 |
-
"loss": 0.
|
| 424 |
"step": 5800
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 4.071773636991028,
|
| 428 |
-
"grad_norm": 0.
|
| 429 |
"learning_rate": 9.67688584689013e-05,
|
| 430 |
-
"loss": 0.
|
| 431 |
"step": 5900
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 4.140786749482402,
|
| 435 |
-
"grad_norm": 0.
|
| 436 |
"learning_rate": 9.655248313910406e-05,
|
| 437 |
-
"loss": 0.
|
| 438 |
"step": 6000
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 4.140786749482402,
|
| 442 |
-
"eval_loss": 0.
|
| 443 |
-
"eval_runtime":
|
| 444 |
-
"eval_samples_per_second":
|
| 445 |
-
"eval_steps_per_second":
|
| 446 |
"step": 6000
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 4.209799861973775,
|
| 450 |
-
"grad_norm": 0.
|
| 451 |
"learning_rate": 9.63293538947056e-05,
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 6100
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 4.278812974465148,
|
| 457 |
-
"grad_norm": 0.
|
| 458 |
"learning_rate": 9.609950310768186e-05,
|
| 459 |
-
"loss": 0.
|
| 460 |
"step": 6200
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 4.3478260869565215,
|
| 464 |
-
"grad_norm": 0.
|
| 465 |
"learning_rate": 9.586296412518169e-05,
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 6300
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 4.416839199447895,
|
| 471 |
-
"grad_norm": 0.
|
| 472 |
"learning_rate": 9.561977126468895e-05,
|
| 473 |
-
"loss": 0.
|
| 474 |
"step": 6400
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 4.485852311939269,
|
| 478 |
-
"grad_norm": 0.
|
| 479 |
"learning_rate": 9.536995980904358e-05,
|
| 480 |
-
"loss": 0.
|
| 481 |
"step": 6500
|
| 482 |
},
|
| 483 |
{
|
| 484 |
"epoch": 4.554865424430642,
|
| 485 |
-
"grad_norm": 0.
|
| 486 |
"learning_rate": 9.511356600132268e-05,
|
| 487 |
-
"loss": 0.
|
| 488 |
"step": 6600
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 4.623878536922016,
|
| 492 |
-
"grad_norm": 0.
|
| 493 |
"learning_rate": 9.485062703958248e-05,
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 6700
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 4.692891649413388,
|
| 499 |
-
"grad_norm": 0.
|
| 500 |
"learning_rate": 9.45811810714613e-05,
|
| 501 |
-
"loss": 0.
|
| 502 |
"step": 6800
|
| 503 |
},
|
| 504 |
{
|
| 505 |
"epoch": 4.761904761904762,
|
| 506 |
-
"grad_norm": 0.
|
| 507 |
"learning_rate": 9.430526718864529e-05,
|
| 508 |
-
"loss": 0.
|
| 509 |
"step": 6900
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 4.830917874396135,
|
| 513 |
-
"grad_norm": 0.
|
| 514 |
"learning_rate": 9.402292542119675e-05,
|
| 515 |
-
"loss": 0.
|
| 516 |
"step": 7000
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 4.899930986887509,
|
| 520 |
-
"grad_norm": 0.
|
| 521 |
"learning_rate": 9.373419673174664e-05,
|
| 522 |
-
"loss": 0.
|
| 523 |
"step": 7100
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 4.9689440993788825,
|
| 527 |
-
"grad_norm": 0.
|
| 528 |
"learning_rate": 9.343912300955152e-05,
|
| 529 |
-
"loss": 0.
|
| 530 |
"step": 7200
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 5.037957211870255,
|
| 534 |
-
"grad_norm": 0.
|
| 535 |
"learning_rate": 9.313774706441632e-05,
|
| 536 |
-
"loss": 0.
|
| 537 |
"step": 7300
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 5.106970324361629,
|
| 541 |
-
"grad_norm": 0.
|
| 542 |
"learning_rate": 9.283011262048334e-05,
|
| 543 |
-
"loss": 0.
|
| 544 |
"step": 7400
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 5.175983436853002,
|
| 548 |
-
"grad_norm": 0.
|
| 549 |
"learning_rate": 9.25162643098886e-05,
|
| 550 |
-
"loss": 0.
|
| 551 |
"step": 7500
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 5.244996549344376,
|
| 555 |
-
"grad_norm": 0.
|
| 556 |
"learning_rate": 9.219624766628675e-05,
|
| 557 |
-
"loss": 0.
|
| 558 |
"step": 7600
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 5.314009661835748,
|
| 562 |
-
"grad_norm": 0.
|
| 563 |
"learning_rate": 9.187010911824475e-05,
|
| 564 |
-
"loss": 0.
|
| 565 |
"step": 7700
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 5.383022774327122,
|
| 569 |
-
"grad_norm": 0.
|
| 570 |
"learning_rate": 9.153789598250606e-05,
|
| 571 |
-
"loss": 0.
|
| 572 |
"step": 7800
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 5.4520358868184955,
|
| 576 |
-
"grad_norm": 0.
|
| 577 |
"learning_rate": 9.119965645712579e-05,
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 7900
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 5.521048999309869,
|
| 583 |
-
"grad_norm": 0.
|
| 584 |
"learning_rate": 9.085543961447808e-05,
|
| 585 |
-
"loss": 0.
|
| 586 |
"step": 8000
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 5.590062111801243,
|
| 590 |
-
"grad_norm": 0.
|
| 591 |
"learning_rate": 9.050529539413655e-05,
|
| 592 |
-
"loss": 0.
|
| 593 |
"step": 8100
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 5.659075224292615,
|
| 597 |
-
"grad_norm": 0.
|
| 598 |
"learning_rate": 9.014927459562901e-05,
|
| 599 |
-
"loss": 0.
|
| 600 |
"step": 8200
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 5.728088336783989,
|
| 604 |
-
"grad_norm": 0.
|
| 605 |
"learning_rate": 8.978742887106742e-05,
|
| 606 |
-
"loss": 0.
|
| 607 |
"step": 8300
|
| 608 |
},
|
| 609 |
{
|
| 610 |
"epoch": 5.797101449275362,
|
| 611 |
-
"grad_norm": 0.
|
| 612 |
"learning_rate": 8.941981071765397e-05,
|
| 613 |
-
"loss": 0.
|
| 614 |
"step": 8400
|
| 615 |
},
|
| 616 |
{
|
| 617 |
"epoch": 5.866114561766736,
|
| 618 |
-
"grad_norm": 0.
|
| 619 |
"learning_rate": 8.904647347006491e-05,
|
| 620 |
-
"loss": 0.
|
| 621 |
"step": 8500
|
| 622 |
},
|
| 623 |
{
|
| 624 |
"epoch": 5.935127674258109,
|
| 625 |
-
"grad_norm": 0.
|
| 626 |
"learning_rate": 8.866747129271244e-05,
|
| 627 |
-
"loss": 0.
|
| 628 |
"step": 8600
|
| 629 |
},
|
| 630 |
{
|
| 631 |
"epoch": 6.004140786749482,
|
| 632 |
-
"grad_norm": 0.
|
| 633 |
"learning_rate": 8.828285917188661e-05,
|
| 634 |
-
"loss": 0.
|
| 635 |
"step": 8700
|
| 636 |
},
|
| 637 |
{
|
| 638 |
"epoch": 6.073153899240856,
|
| 639 |
-
"grad_norm": 0.
|
| 640 |
"learning_rate": 8.789269290777772e-05,
|
| 641 |
-
"loss": 0.
|
| 642 |
"step": 8800
|
| 643 |
},
|
| 644 |
{
|
| 645 |
"epoch": 6.142167011732229,
|
| 646 |
-
"grad_norm": 0.
|
| 647 |
"learning_rate": 8.74970291063808e-05,
|
| 648 |
-
"loss": 0.
|
| 649 |
"step": 8900
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"epoch": 6.211180124223603,
|
| 653 |
-
"grad_norm": 0.
|
| 654 |
"learning_rate": 8.709592517128303e-05,
|
| 655 |
-
"loss": 0.
|
| 656 |
"step": 9000
|
| 657 |
},
|
| 658 |
{
|
| 659 |
"epoch": 6.211180124223603,
|
| 660 |
-
"eval_loss": 0.
|
| 661 |
-
"eval_runtime":
|
| 662 |
-
"eval_samples_per_second":
|
| 663 |
-
"eval_steps_per_second":
|
| 664 |
"step": 9000
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 6.280193236714976,
|
| 668 |
-
"grad_norm":
|
| 669 |
"learning_rate": 8.668943929533559e-05,
|
| 670 |
-
"loss": 0.
|
| 671 |
"step": 9100
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 6.349206349206349,
|
| 675 |
-
"grad_norm": 0.
|
| 676 |
"learning_rate": 8.627763045221097e-05,
|
| 677 |
-
"loss": 0.
|
| 678 |
"step": 9200
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 6.418219461697722,
|
| 682 |
-
"grad_norm": 0.
|
| 683 |
"learning_rate": 8.586055838784687e-05,
|
| 684 |
-
"loss": 0.
|
| 685 |
"step": 9300
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 6.487232574189096,
|
| 689 |
-
"grad_norm": 0.
|
| 690 |
"learning_rate": 8.543828361177825e-05,
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 9400
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 6.5562456866804695,
|
| 696 |
-
"grad_norm": 0.
|
| 697 |
"learning_rate": 8.501086738835843e-05,
|
| 698 |
-
"loss": 0.
|
| 699 |
"step": 9500
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 6.625258799171843,
|
| 703 |
-
"grad_norm": 0.
|
| 704 |
"learning_rate": 8.457837172787082e-05,
|
| 705 |
-
"loss": 0.
|
| 706 |
"step": 9600
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 6.694271911663216,
|
| 710 |
-
"grad_norm": 0.
|
| 711 |
"learning_rate": 8.414085937753229e-05,
|
| 712 |
-
"loss": 0.
|
| 713 |
"step": 9700
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 6.763285024154589,
|
| 717 |
-
"grad_norm": 0.
|
| 718 |
"learning_rate": 8.369839381238977e-05,
|
| 719 |
-
"loss": 0.
|
| 720 |
"step": 9800
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 6.832298136645963,
|
| 724 |
-
"grad_norm": 0.
|
| 725 |
"learning_rate": 8.325103922611114e-05,
|
| 726 |
-
"loss": 0.
|
| 727 |
"step": 9900
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 6.901311249137336,
|
| 731 |
-
"grad_norm": 0.
|
| 732 |
"learning_rate": 8.279886052167185e-05,
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 10000
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 6.97032436162871,
|
| 738 |
-
"grad_norm": 0.
|
| 739 |
"learning_rate": 8.234192330193881e-05,
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 10100
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 7.0393374741200825,
|
| 745 |
-
"grad_norm": 0.
|
| 746 |
"learning_rate": 8.18802938601525e-05,
|
| 747 |
-
"loss": 0.
|
| 748 |
"step": 10200
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 7.108350586611456,
|
| 752 |
-
"grad_norm": 0.
|
| 753 |
"learning_rate": 8.14140391703091e-05,
|
| 754 |
-
"loss": 0.
|
| 755 |
"step": 10300
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 7.17736369910283,
|
| 759 |
-
"grad_norm": 0.
|
| 760 |
"learning_rate": 8.094322687744371e-05,
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 10400
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 7.246376811594203,
|
| 766 |
-
"grad_norm": 0.
|
| 767 |
"learning_rate": 8.046792528781642e-05,
|
| 768 |
-
"loss": 0.
|
| 769 |
"step": 10500
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 7.315389924085577,
|
| 773 |
-
"grad_norm": 0.
|
| 774 |
"learning_rate": 7.998820335900217e-05,
|
| 775 |
-
"loss": 0.
|
| 776 |
"step": 10600
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 7.384403036576949,
|
| 780 |
-
"grad_norm": 0.
|
| 781 |
"learning_rate": 7.950413068988638e-05,
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 10700
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 7.453416149068323,
|
| 787 |
-
"grad_norm": 0.
|
| 788 |
"learning_rate": 7.901577751056739e-05,
|
| 789 |
-
"loss": 0.
|
| 790 |
"step": 10800
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 7.522429261559696,
|
| 794 |
-
"grad_norm": 0.
|
| 795 |
"learning_rate": 7.85232146721673e-05,
|
| 796 |
-
"loss": 0.
|
| 797 |
"step": 10900
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 7.59144237405107,
|
| 801 |
-
"grad_norm": 0.
|
| 802 |
"learning_rate": 7.802651363655285e-05,
|
| 803 |
-
"loss": 0.
|
| 804 |
"step": 11000
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 7.660455486542443,
|
| 808 |
-
"grad_norm": 0.
|
| 809 |
"learning_rate": 7.75257464659676e-05,
|
| 810 |
-
"loss": 0.
|
| 811 |
"step": 11100
|
| 812 |
},
|
| 813 |
{
|
| 814 |
"epoch": 7.729468599033816,
|
| 815 |
-
"grad_norm": 0.
|
| 816 |
"learning_rate": 7.702098581257692e-05,
|
| 817 |
-
"loss": 0.
|
| 818 |
"step": 11200
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 7.79848171152519,
|
| 822 |
-
"grad_norm": 0.
|
| 823 |
"learning_rate": 7.651230490792762e-05,
|
| 824 |
-
"loss": 0.
|
| 825 |
"step": 11300
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 7.867494824016563,
|
| 829 |
-
"grad_norm": 0.
|
| 830 |
"learning_rate": 7.599977755232328e-05,
|
| 831 |
-
"loss": 0.
|
| 832 |
"step": 11400
|
| 833 |
},
|
| 834 |
{
|
| 835 |
"epoch": 7.936507936507937,
|
| 836 |
-
"grad_norm": 0.
|
| 837 |
"learning_rate": 7.548347810411729e-05,
|
| 838 |
-
"loss": 0.
|
| 839 |
"step": 11500
|
| 840 |
},
|
| 841 |
{
|
| 842 |
"epoch": 8.00552104899931,
|
| 843 |
-
"grad_norm": 0.
|
| 844 |
"learning_rate": 7.496348146892469e-05,
|
| 845 |
-
"loss": 0.
|
| 846 |
"step": 11600
|
| 847 |
},
|
| 848 |
{
|
| 849 |
"epoch": 8.074534161490684,
|
| 850 |
-
"grad_norm": 0.
|
| 851 |
"learning_rate": 7.443986308875479e-05,
|
| 852 |
-
"loss": 0.
|
| 853 |
"step": 11700
|
| 854 |
},
|
| 855 |
{
|
| 856 |
"epoch": 8.143547273982056,
|
| 857 |
-
"grad_norm": 0.
|
| 858 |
"learning_rate": 7.391269893106592e-05,
|
| 859 |
-
"loss": 0.
|
| 860 |
"step": 11800
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 8.21256038647343,
|
| 864 |
-
"grad_norm": 0.
|
| 865 |
"learning_rate": 7.338206547774395e-05,
|
| 866 |
-
"loss": 0.
|
| 867 |
"step": 11900
|
| 868 |
},
|
| 869 |
{
|
| 870 |
"epoch": 8.281573498964804,
|
| 871 |
-
"grad_norm": 0.
|
| 872 |
"learning_rate": 7.284803971400614e-05,
|
| 873 |
-
"loss": 0.
|
| 874 |
"step": 12000
|
| 875 |
},
|
| 876 |
{
|
| 877 |
"epoch": 8.281573498964804,
|
| 878 |
-
"eval_loss": 0.
|
| 879 |
-
"eval_runtime":
|
| 880 |
-
"eval_samples_per_second":
|
| 881 |
-
"eval_steps_per_second":
|
| 882 |
"step": 12000
|
| 883 |
}
|
| 884 |
],
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.06901311249137336,
|
| 14 |
+
"grad_norm": 1.70753812789917,
|
| 15 |
"learning_rate": 3.4161490683229816e-06,
|
| 16 |
+
"loss": 0.6597207641601562,
|
| 17 |
"step": 100
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.13802622498274672,
|
| 21 |
+
"grad_norm": 0.6806356906890869,
|
| 22 |
"learning_rate": 6.86680469289165e-06,
|
| 23 |
+
"loss": 0.2926066780090332,
|
| 24 |
"step": 200
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.2070393374741201,
|
| 28 |
+
"grad_norm": 1.1896523237228394,
|
| 29 |
"learning_rate": 1.0317460317460318e-05,
|
| 30 |
+
"loss": 0.184893798828125,
|
| 31 |
"step": 300
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.27605244996549344,
|
| 35 |
+
"grad_norm": 1.0565272569656372,
|
| 36 |
"learning_rate": 1.3768115942028985e-05,
|
| 37 |
+
"loss": 0.16684833526611328,
|
| 38 |
"step": 400
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.3450655624568668,
|
| 42 |
+
"grad_norm": 1.4511226415634155,
|
| 43 |
"learning_rate": 1.7218771566597654e-05,
|
| 44 |
+
"loss": 0.15068528175354004,
|
| 45 |
"step": 500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.4140786749482402,
|
| 49 |
+
"grad_norm": 0.8944125771522522,
|
| 50 |
"learning_rate": 2.0669427191166323e-05,
|
| 51 |
+
"loss": 0.1340955924987793,
|
| 52 |
"step": 600
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.4830917874396135,
|
| 56 |
+
"grad_norm": 1.0738117694854736,
|
| 57 |
"learning_rate": 2.412008281573499e-05,
|
| 58 |
+
"loss": 0.1130269718170166,
|
| 59 |
"step": 700
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.5521048999309869,
|
| 63 |
+
"grad_norm": 0.7529744505882263,
|
| 64 |
"learning_rate": 2.757073844030366e-05,
|
| 65 |
+
"loss": 0.10459966659545898,
|
| 66 |
"step": 800
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.6211180124223602,
|
| 70 |
+
"grad_norm": 1.183493733406067,
|
| 71 |
"learning_rate": 3.102139406487233e-05,
|
| 72 |
+
"loss": 0.09097274780273437,
|
| 73 |
"step": 900
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.6901311249137336,
|
| 77 |
+
"grad_norm": 0.8895713686943054,
|
| 78 |
"learning_rate": 3.4472049689440996e-05,
|
| 79 |
+
"loss": 0.0841109275817871,
|
| 80 |
"step": 1000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.759144237405107,
|
| 84 |
+
"grad_norm": 0.8567629456520081,
|
| 85 |
"learning_rate": 3.7922705314009665e-05,
|
| 86 |
+
"loss": 0.07087889671325684,
|
| 87 |
"step": 1100
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.8281573498964804,
|
| 91 |
+
"grad_norm": 0.8626934885978699,
|
| 92 |
"learning_rate": 4.1373360938578333e-05,
|
| 93 |
+
"loss": 0.059555401802062986,
|
| 94 |
"step": 1200
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.8971704623878537,
|
| 98 |
+
"grad_norm": 0.5157253742218018,
|
| 99 |
"learning_rate": 4.4824016563146996e-05,
|
| 100 |
+
"loss": 0.059585466384887695,
|
| 101 |
"step": 1300
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.966183574879227,
|
| 105 |
+
"grad_norm": 0.9269208312034607,
|
| 106 |
"learning_rate": 4.827467218771567e-05,
|
| 107 |
+
"loss": 0.05334312915802002,
|
| 108 |
"step": 1400
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 1.0351966873706004,
|
| 112 |
+
"grad_norm": 1.4995015859603882,
|
| 113 |
"learning_rate": 5.172532781228434e-05,
|
| 114 |
+
"loss": 0.04970784664154053,
|
| 115 |
"step": 1500
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 1.1042097998619738,
|
| 119 |
+
"grad_norm": 0.6324323415756226,
|
| 120 |
"learning_rate": 5.517598343685301e-05,
|
| 121 |
+
"loss": 0.04526669979095459,
|
| 122 |
"step": 1600
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 1.173222912353347,
|
| 126 |
+
"grad_norm": 1.5328924655914307,
|
| 127 |
"learning_rate": 5.862663906142167e-05,
|
| 128 |
+
"loss": 0.043799943923950195,
|
| 129 |
"step": 1700
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 1.2422360248447206,
|
| 133 |
+
"grad_norm": 1.0417181253433228,
|
| 134 |
"learning_rate": 6.207729468599034e-05,
|
| 135 |
+
"loss": 0.038412556648254395,
|
| 136 |
"step": 1800
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 1.311249137336094,
|
| 140 |
+
"grad_norm": 0.7594957947731018,
|
| 141 |
"learning_rate": 6.5527950310559e-05,
|
| 142 |
+
"loss": 0.03864224910736084,
|
| 143 |
"step": 1900
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 1.3802622498274673,
|
| 147 |
+
"grad_norm": 0.7584347128868103,
|
| 148 |
"learning_rate": 6.897860593512768e-05,
|
| 149 |
+
"loss": 0.037553794384002685,
|
| 150 |
"step": 2000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 1.4492753623188406,
|
| 154 |
+
"grad_norm": 1.3384641408920288,
|
| 155 |
"learning_rate": 7.242926155969635e-05,
|
| 156 |
+
"loss": 0.035451841354370114,
|
| 157 |
"step": 2100
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 1.518288474810214,
|
| 161 |
+
"grad_norm": 1.9227489233016968,
|
| 162 |
"learning_rate": 7.587991718426502e-05,
|
| 163 |
+
"loss": 0.032987830638885496,
|
| 164 |
"step": 2200
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 1.5873015873015874,
|
| 168 |
+
"grad_norm": 0.33554157614707947,
|
| 169 |
"learning_rate": 7.933057280883368e-05,
|
| 170 |
+
"loss": 0.03358668804168701,
|
| 171 |
"step": 2300
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 1.6563146997929605,
|
| 175 |
+
"grad_norm": 1.8089464902877808,
|
| 176 |
"learning_rate": 8.278122843340234e-05,
|
| 177 |
+
"loss": 0.030594797134399415,
|
| 178 |
"step": 2400
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 1.725327812284334,
|
| 182 |
+
"grad_norm": 2.194699764251709,
|
| 183 |
"learning_rate": 8.623188405797103e-05,
|
| 184 |
+
"loss": 0.02913449764251709,
|
| 185 |
"step": 2500
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 1.7943409247757074,
|
| 189 |
+
"grad_norm": 0.5577589869499207,
|
| 190 |
"learning_rate": 8.968253968253969e-05,
|
| 191 |
+
"loss": 0.02450634479522705,
|
| 192 |
"step": 2600
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 1.8633540372670807,
|
| 196 |
+
"grad_norm": 0.9141516089439392,
|
| 197 |
"learning_rate": 9.313319530710835e-05,
|
| 198 |
+
"loss": 0.026773200035095215,
|
| 199 |
"step": 2700
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 1.9323671497584543,
|
| 203 |
+
"grad_norm": 0.19354890286922455,
|
| 204 |
"learning_rate": 9.658385093167702e-05,
|
| 205 |
+
"loss": 0.025803778171539307,
|
| 206 |
"step": 2800
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 2.0013802622498273,
|
| 210 |
+
"grad_norm": 0.23801928758621216,
|
| 211 |
"learning_rate": 9.99999996372913e-05,
|
| 212 |
+
"loss": 0.023492772579193116,
|
| 213 |
"step": 2900
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 2.070393374741201,
|
| 217 |
+
"grad_norm": 0.32787156105041504,
|
| 218 |
"learning_rate": 9.999630005418305e-05,
|
| 219 |
+
"loss": 0.019941968917846678,
|
| 220 |
"step": 3000
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 2.070393374741201,
|
| 224 |
+
"eval_loss": 0.03612843528389931,
|
| 225 |
+
"eval_runtime": 239.8677,
|
| 226 |
+
"eval_samples_per_second": 23.467,
|
| 227 |
+
"eval_steps_per_second": 5.87,
|
| 228 |
"step": 3000
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 2.139406487232574,
|
| 232 |
+
"grad_norm": 0.5818076729774475,
|
| 233 |
"learning_rate": 9.998534692157168e-05,
|
| 234 |
+
"loss": 0.02141228437423706,
|
| 235 |
"step": 3100
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 2.2084195997239475,
|
| 239 |
+
"grad_norm": 0.3472735583782196,
|
| 240 |
"learning_rate": 9.99671418285566e-05,
|
| 241 |
+
"loss": 0.01901582717895508,
|
| 242 |
"step": 3200
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 2.277432712215321,
|
| 246 |
+
"grad_norm": 0.4493163228034973,
|
| 247 |
"learning_rate": 9.994168741636414e-05,
|
| 248 |
+
"loss": 0.018842363357543947,
|
| 249 |
"step": 3300
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 2.346445824706694,
|
| 253 |
+
"grad_norm": 0.4030870497226715,
|
| 254 |
"learning_rate": 9.990898737796431e-05,
|
| 255 |
+
"loss": 0.02044015645980835,
|
| 256 |
"step": 3400
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 2.4154589371980677,
|
| 260 |
+
"grad_norm": 0.3151642978191376,
|
| 261 |
"learning_rate": 9.986904645753517e-05,
|
| 262 |
+
"loss": 0.017371728420257568,
|
| 263 |
"step": 3500
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 2.4844720496894412,
|
| 267 |
+
"grad_norm": 0.2812744379043579,
|
| 268 |
"learning_rate": 9.982187044977433e-05,
|
| 269 |
+
"loss": 0.0203560471534729,
|
| 270 |
"step": 3600
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 2.5534851621808143,
|
| 274 |
+
"grad_norm": 0.3646181523799896,
|
| 275 |
"learning_rate": 9.976746619905848e-05,
|
| 276 |
+
"loss": 0.016241873502731322,
|
| 277 |
"step": 3700
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 2.622498274672188,
|
| 281 |
+
"grad_norm": 0.16350792348384857,
|
| 282 |
"learning_rate": 9.970584159845019e-05,
|
| 283 |
+
"loss": 0.016487326622009277,
|
| 284 |
"step": 3800
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 2.691511387163561,
|
| 288 |
+
"grad_norm": 0.32364580035209656,
|
| 289 |
"learning_rate": 9.963700558855286e-05,
|
| 290 |
+
"loss": 0.014688656330108643,
|
| 291 |
"step": 3900
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 2.7605244996549345,
|
| 295 |
+
"grad_norm": 0.9933033585548401,
|
| 296 |
"learning_rate": 9.956096815621365e-05,
|
| 297 |
+
"loss": 0.015503013134002685,
|
| 298 |
"step": 4000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 2.829537612146308,
|
| 302 |
+
"grad_norm": 0.20645038783550262,
|
| 303 |
"learning_rate": 9.947774033307444e-05,
|
| 304 |
+
"loss": 0.013945527076721191,
|
| 305 |
"step": 4100
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 2.898550724637681,
|
| 309 |
+
"grad_norm": 0.2992095649242401,
|
| 310 |
"learning_rate": 9.938733419397148e-05,
|
| 311 |
+
"loss": 0.01564746379852295,
|
| 312 |
"step": 4200
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 2.9675638371290542,
|
| 316 |
+
"grad_norm": 0.15112999081611633,
|
| 317 |
"learning_rate": 9.928976285518347e-05,
|
| 318 |
+
"loss": 0.014093859195709229,
|
| 319 |
"step": 4300
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 3.036576949620428,
|
| 323 |
+
"grad_norm": 0.21860471367835999,
|
| 324 |
"learning_rate": 9.918504047252866e-05,
|
| 325 |
+
"loss": 0.013013427257537841,
|
| 326 |
"step": 4400
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 3.1055900621118013,
|
| 330 |
+
"grad_norm": 0.8001585006713867,
|
| 331 |
"learning_rate": 9.907318223931104e-05,
|
| 332 |
+
"loss": 0.010665929317474366,
|
| 333 |
"step": 4500
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 3.1746031746031744,
|
| 337 |
+
"grad_norm": 0.2942330241203308,
|
| 338 |
"learning_rate": 9.895420438411616e-05,
|
| 339 |
+
"loss": 0.01030845046043396,
|
| 340 |
"step": 4600
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 3.243616287094548,
|
| 344 |
+
"grad_norm": 0.4483569860458374,
|
| 345 |
"learning_rate": 9.882812416845659e-05,
|
| 346 |
+
"loss": 0.012794067859649658,
|
| 347 |
"step": 4700
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 3.3126293995859215,
|
| 351 |
+
"grad_norm": 0.10391439497470856,
|
| 352 |
"learning_rate": 9.869495988426765e-05,
|
| 353 |
+
"loss": 0.010797868967056274,
|
| 354 |
"step": 4800
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 3.3816425120772946,
|
| 358 |
+
"grad_norm": 0.17631374299526215,
|
| 359 |
"learning_rate": 9.85547308512535e-05,
|
| 360 |
+
"loss": 0.012075445652008056,
|
| 361 |
"step": 4900
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 3.450655624568668,
|
| 365 |
+
"grad_norm": 0.5997236371040344,
|
| 366 |
"learning_rate": 9.84074574140843e-05,
|
| 367 |
+
"loss": 0.009871997833251954,
|
| 368 |
"step": 5000
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 3.5196687370600412,
|
| 372 |
+
"grad_norm": 0.6598154306411743,
|
| 373 |
"learning_rate": 9.825316093944449e-05,
|
| 374 |
+
"loss": 0.009196481704711913,
|
| 375 |
"step": 5100
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 3.588681849551415,
|
| 379 |
+
"grad_norm": 0.3595139980316162,
|
| 380 |
"learning_rate": 9.809186381293293e-05,
|
| 381 |
+
"loss": 0.011774812936782836,
|
| 382 |
"step": 5200
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 3.657694962042788,
|
| 386 |
+
"grad_norm": 0.33344539999961853,
|
| 387 |
"learning_rate": 9.792358943581512e-05,
|
| 388 |
+
"loss": 0.009864444136619568,
|
| 389 |
"step": 5300
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 3.7267080745341614,
|
| 393 |
+
"grad_norm": 0.4517774283885956,
|
| 394 |
"learning_rate": 9.774836222162811e-05,
|
| 395 |
+
"loss": 0.009815619587898254,
|
| 396 |
"step": 5400
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 3.795721187025535,
|
| 400 |
+
"grad_norm": 0.3263384997844696,
|
| 401 |
"learning_rate": 9.756620759263859e-05,
|
| 402 |
+
"loss": 0.008986613750457763,
|
| 403 |
"step": 5500
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 3.864734299516908,
|
| 407 |
+
"grad_norm": 0.1386384516954422,
|
| 408 |
"learning_rate": 9.73771519761545e-05,
|
| 409 |
+
"loss": 0.008247822523117065,
|
| 410 |
"step": 5600
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 3.9337474120082816,
|
| 414 |
+
"grad_norm": 0.3077174127101898,
|
| 415 |
"learning_rate": 9.718122280069101e-05,
|
| 416 |
+
"loss": 0.00787350833415985,
|
| 417 |
"step": 5700
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 4.002760524499655,
|
| 421 |
+
"grad_norm": 0.15773248672485352,
|
| 422 |
"learning_rate": 9.697844849199104e-05,
|
| 423 |
+
"loss": 0.00808551549911499,
|
| 424 |
"step": 5800
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 4.071773636991028,
|
| 428 |
+
"grad_norm": 0.6109098792076111,
|
| 429 |
"learning_rate": 9.67688584689013e-05,
|
| 430 |
+
"loss": 0.0070124870538711545,
|
| 431 |
"step": 5900
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 4.140786749482402,
|
| 435 |
+
"grad_norm": 0.12046466022729874,
|
| 436 |
"learning_rate": 9.655248313910406e-05,
|
| 437 |
+
"loss": 0.00716442346572876,
|
| 438 |
"step": 6000
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 4.140786749482402,
|
| 442 |
+
"eval_loss": 0.02697061188519001,
|
| 443 |
+
"eval_runtime": 239.587,
|
| 444 |
+
"eval_samples_per_second": 23.495,
|
| 445 |
+
"eval_steps_per_second": 5.877,
|
| 446 |
"step": 6000
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 4.209799861973775,
|
| 450 |
+
"grad_norm": 0.07307147979736328,
|
| 451 |
"learning_rate": 9.63293538947056e-05,
|
| 452 |
+
"loss": 0.005638118982315064,
|
| 453 |
"step": 6100
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 4.278812974465148,
|
| 457 |
+
"grad_norm": 0.16364820301532745,
|
| 458 |
"learning_rate": 9.609950310768186e-05,
|
| 459 |
+
"loss": 0.007392752766609192,
|
| 460 |
"step": 6200
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 4.3478260869565215,
|
| 464 |
+
"grad_norm": 0.26814261078834534,
|
| 465 |
"learning_rate": 9.586296412518169e-05,
|
| 466 |
+
"loss": 0.005863593816757202,
|
| 467 |
"step": 6300
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 4.416839199447895,
|
| 471 |
+
"grad_norm": 0.04622579738497734,
|
| 472 |
"learning_rate": 9.561977126468895e-05,
|
| 473 |
+
"loss": 0.007532884478569031,
|
| 474 |
"step": 6400
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 4.485852311939269,
|
| 478 |
+
"grad_norm": 0.4954683184623718,
|
| 479 |
"learning_rate": 9.536995980904358e-05,
|
| 480 |
+
"loss": 0.007324672937393188,
|
| 481 |
"step": 6500
|
| 482 |
},
|
| 483 |
{
|
| 484 |
"epoch": 4.554865424430642,
|
| 485 |
+
"grad_norm": 0.15553529560565948,
|
| 486 |
"learning_rate": 9.511356600132268e-05,
|
| 487 |
+
"loss": 0.006730241775512695,
|
| 488 |
"step": 6600
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 4.623878536922016,
|
| 492 |
+
"grad_norm": 0.011446000076830387,
|
| 493 |
"learning_rate": 9.485062703958248e-05,
|
| 494 |
+
"loss": 0.005915153026580811,
|
| 495 |
"step": 6700
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 4.692891649413388,
|
| 499 |
+
"grad_norm": 0.19122876226902008,
|
| 500 |
"learning_rate": 9.45811810714613e-05,
|
| 501 |
+
"loss": 0.006628613471984863,
|
| 502 |
"step": 6800
|
| 503 |
},
|
| 504 |
{
|
| 505 |
"epoch": 4.761904761904762,
|
| 506 |
+
"grad_norm": 0.10010533034801483,
|
| 507 |
"learning_rate": 9.430526718864529e-05,
|
| 508 |
+
"loss": 0.0048975437879562374,
|
| 509 |
"step": 6900
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 4.830917874396135,
|
| 513 |
+
"grad_norm": 0.01408712100237608,
|
| 514 |
"learning_rate": 9.402292542119675e-05,
|
| 515 |
+
"loss": 0.006436612606048584,
|
| 516 |
"step": 7000
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 4.899930986887509,
|
| 520 |
+
"grad_norm": 0.16825127601623535,
|
| 521 |
"learning_rate": 9.373419673174664e-05,
|
| 522 |
+
"loss": 0.00532124400138855,
|
| 523 |
"step": 7100
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 4.9689440993788825,
|
| 527 |
+
"grad_norm": 0.15378710627555847,
|
| 528 |
"learning_rate": 9.343912300955152e-05,
|
| 529 |
+
"loss": 0.007193892002105713,
|
| 530 |
"step": 7200
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 5.037957211870255,
|
| 534 |
+
"grad_norm": 0.021719099953770638,
|
| 535 |
"learning_rate": 9.313774706441632e-05,
|
| 536 |
+
"loss": 0.004890411496162415,
|
| 537 |
"step": 7300
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 5.106970324361629,
|
| 541 |
+
"grad_norm": 0.23345591127872467,
|
| 542 |
"learning_rate": 9.283011262048334e-05,
|
| 543 |
+
"loss": 0.004248523116111755,
|
| 544 |
"step": 7400
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 5.175983436853002,
|
| 548 |
+
"grad_norm": 0.05435383692383766,
|
| 549 |
"learning_rate": 9.25162643098886e-05,
|
| 550 |
+
"loss": 0.004317086935043335,
|
| 551 |
"step": 7500
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 5.244996549344376,
|
| 555 |
+
"grad_norm": 0.5512757897377014,
|
| 556 |
"learning_rate": 9.219624766628675e-05,
|
| 557 |
+
"loss": 0.004326205849647522,
|
| 558 |
"step": 7600
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 5.314009661835748,
|
| 562 |
+
"grad_norm": 0.02267337404191494,
|
| 563 |
"learning_rate": 9.187010911824475e-05,
|
| 564 |
+
"loss": 0.005061171054840088,
|
| 565 |
"step": 7700
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 5.383022774327122,
|
| 569 |
+
"grad_norm": 0.6639739274978638,
|
| 570 |
"learning_rate": 9.153789598250606e-05,
|
| 571 |
+
"loss": 0.0038250452280044556,
|
| 572 |
"step": 7800
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 5.4520358868184955,
|
| 576 |
+
"grad_norm": 0.043474406003952026,
|
| 577 |
"learning_rate": 9.119965645712579e-05,
|
| 578 |
+
"loss": 0.005064901709556579,
|
| 579 |
"step": 7900
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 5.521048999309869,
|
| 583 |
+
"grad_norm": 0.05946451053023338,
|
| 584 |
"learning_rate": 9.085543961447808e-05,
|
| 585 |
+
"loss": 0.005857201814651489,
|
| 586 |
"step": 8000
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 5.590062111801243,
|
| 590 |
+
"grad_norm": 0.20854756236076355,
|
| 591 |
"learning_rate": 9.050529539413655e-05,
|
| 592 |
+
"loss": 0.0038475406169891356,
|
| 593 |
"step": 8100
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 5.659075224292615,
|
| 597 |
+
"grad_norm": 0.2599511742591858,
|
| 598 |
"learning_rate": 9.014927459562901e-05,
|
| 599 |
+
"loss": 0.004083446562290192,
|
| 600 |
"step": 8200
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 5.728088336783989,
|
| 604 |
+
"grad_norm": 0.116208516061306,
|
| 605 |
"learning_rate": 8.978742887106742e-05,
|
| 606 |
+
"loss": 0.005109227299690247,
|
| 607 |
"step": 8300
|
| 608 |
},
|
| 609 |
{
|
| 610 |
"epoch": 5.797101449275362,
|
| 611 |
+
"grad_norm": 0.08974844962358475,
|
| 612 |
"learning_rate": 8.941981071765397e-05,
|
| 613 |
+
"loss": 0.006279203295707703,
|
| 614 |
"step": 8400
|
| 615 |
},
|
| 616 |
{
|
| 617 |
"epoch": 5.866114561766736,
|
| 618 |
+
"grad_norm": 0.7637934684753418,
|
| 619 |
"learning_rate": 8.904647347006491e-05,
|
| 620 |
+
"loss": 0.004381818473339081,
|
| 621 |
"step": 8500
|
| 622 |
},
|
| 623 |
{
|
| 624 |
"epoch": 5.935127674258109,
|
| 625 |
+
"grad_norm": 0.04625600576400757,
|
| 626 |
"learning_rate": 8.866747129271244e-05,
|
| 627 |
+
"loss": 0.005535225868225097,
|
| 628 |
"step": 8600
|
| 629 |
},
|
| 630 |
{
|
| 631 |
"epoch": 6.004140786749482,
|
| 632 |
+
"grad_norm": 0.248281329870224,
|
| 633 |
"learning_rate": 8.828285917188661e-05,
|
| 634 |
+
"loss": 0.004098294377326965,
|
| 635 |
"step": 8700
|
| 636 |
},
|
| 637 |
{
|
| 638 |
"epoch": 6.073153899240856,
|
| 639 |
+
"grad_norm": 0.19941352307796478,
|
| 640 |
"learning_rate": 8.789269290777772e-05,
|
| 641 |
+
"loss": 0.0029554140567779543,
|
| 642 |
"step": 8800
|
| 643 |
},
|
| 644 |
{
|
| 645 |
"epoch": 6.142167011732229,
|
| 646 |
+
"grad_norm": 0.2032374143600464,
|
| 647 |
"learning_rate": 8.74970291063808e-05,
|
| 648 |
+
"loss": 0.004462797343730927,
|
| 649 |
"step": 8900
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"epoch": 6.211180124223603,
|
| 653 |
+
"grad_norm": 0.0009751737816259265,
|
| 654 |
"learning_rate": 8.709592517128303e-05,
|
| 655 |
+
"loss": 0.0022404642403125764,
|
| 656 |
"step": 9000
|
| 657 |
},
|
| 658 |
{
|
| 659 |
"epoch": 6.211180124223603,
|
| 660 |
+
"eval_loss": 0.03547283634543419,
|
| 661 |
+
"eval_runtime": 239.945,
|
| 662 |
+
"eval_samples_per_second": 23.46,
|
| 663 |
+
"eval_steps_per_second": 5.868,
|
| 664 |
"step": 9000
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 6.280193236714976,
|
| 668 |
+
"grad_norm": 0.3267335295677185,
|
| 669 |
"learning_rate": 8.668943929533559e-05,
|
| 670 |
+
"loss": 0.004250606298446655,
|
| 671 |
"step": 9100
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 6.349206349206349,
|
| 675 |
+
"grad_norm": 0.2648331820964813,
|
| 676 |
"learning_rate": 8.627763045221097e-05,
|
| 677 |
+
"loss": 0.003961024880409241,
|
| 678 |
"step": 9200
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 6.418219461697722,
|
| 682 |
+
"grad_norm": 0.005424227099865675,
|
| 683 |
"learning_rate": 8.586055838784687e-05,
|
| 684 |
+
"loss": 0.0031815102696418763,
|
| 685 |
"step": 9300
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 6.487232574189096,
|
| 689 |
+
"grad_norm": 0.17170420289039612,
|
| 690 |
"learning_rate": 8.543828361177825e-05,
|
| 691 |
+
"loss": 0.002956092357635498,
|
| 692 |
"step": 9400
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 6.5562456866804695,
|
| 696 |
+
"grad_norm": 0.04011496528983116,
|
| 697 |
"learning_rate": 8.501086738835843e-05,
|
| 698 |
+
"loss": 0.001872006058692932,
|
| 699 |
"step": 9500
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 6.625258799171843,
|
| 703 |
+
"grad_norm": 0.38712838292121887,
|
| 704 |
"learning_rate": 8.457837172787082e-05,
|
| 705 |
+
"loss": 0.0033960705995559693,
|
| 706 |
"step": 9600
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 6.694271911663216,
|
| 710 |
+
"grad_norm": 0.02333800308406353,
|
| 711 |
"learning_rate": 8.414085937753229e-05,
|
| 712 |
+
"loss": 0.004199814796447754,
|
| 713 |
"step": 9700
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 6.763285024154589,
|
| 717 |
+
"grad_norm": 0.23009498417377472,
|
| 718 |
"learning_rate": 8.369839381238977e-05,
|
| 719 |
+
"loss": 0.00346614807844162,
|
| 720 |
"step": 9800
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 6.832298136645963,
|
| 724 |
+
"grad_norm": 0.004106150008738041,
|
| 725 |
"learning_rate": 8.325103922611114e-05,
|
| 726 |
+
"loss": 0.002711970806121826,
|
| 727 |
"step": 9900
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 6.901311249137336,
|
| 731 |
+
"grad_norm": 0.12709400057792664,
|
| 732 |
"learning_rate": 8.279886052167185e-05,
|
| 733 |
+
"loss": 0.002297592759132385,
|
| 734 |
"step": 10000
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 6.97032436162871,
|
| 738 |
+
"grad_norm": 0.47119075059890747,
|
| 739 |
"learning_rate": 8.234192330193881e-05,
|
| 740 |
+
"loss": 0.004861023426055908,
|
| 741 |
"step": 10100
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 7.0393374741200825,
|
| 745 |
+
"grad_norm": 0.1427253782749176,
|
| 746 |
"learning_rate": 8.18802938601525e-05,
|
| 747 |
+
"loss": 0.0019143480062484741,
|
| 748 |
"step": 10200
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 7.108350586611456,
|
| 752 |
+
"grad_norm": 0.0031943873036652803,
|
| 753 |
"learning_rate": 8.14140391703091e-05,
|
| 754 |
+
"loss": 0.0019206857681274413,
|
| 755 |
"step": 10300
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 7.17736369910283,
|
| 759 |
+
"grad_norm": 0.004871256183832884,
|
| 760 |
"learning_rate": 8.094322687744371e-05,
|
| 761 |
+
"loss": 0.002094006538391113,
|
| 762 |
"step": 10400
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 7.246376811594203,
|
| 766 |
+
"grad_norm": 0.15592052042484283,
|
| 767 |
"learning_rate": 8.046792528781642e-05,
|
| 768 |
+
"loss": 0.001511605978012085,
|
| 769 |
"step": 10500
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 7.315389924085577,
|
| 773 |
+
"grad_norm": 0.6760014295578003,
|
| 774 |
"learning_rate": 7.998820335900217e-05,
|
| 775 |
+
"loss": 0.002429080605506897,
|
| 776 |
"step": 10600
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 7.384403036576949,
|
| 780 |
+
"grad_norm": 0.014859587885439396,
|
| 781 |
"learning_rate": 7.950413068988638e-05,
|
| 782 |
+
"loss": 0.0028734594583511353,
|
| 783 |
"step": 10700
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 7.453416149068323,
|
| 787 |
+
"grad_norm": 0.03109767660498619,
|
| 788 |
"learning_rate": 7.901577751056739e-05,
|
| 789 |
+
"loss": 0.002481156587600708,
|
| 790 |
"step": 10800
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 7.522429261559696,
|
| 794 |
+
"grad_norm": 0.020552068948745728,
|
| 795 |
"learning_rate": 7.85232146721673e-05,
|
| 796 |
+
"loss": 0.002704034745693207,
|
| 797 |
"step": 10900
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 7.59144237405107,
|
| 801 |
+
"grad_norm": 0.8788889050483704,
|
| 802 |
"learning_rate": 7.802651363655285e-05,
|
| 803 |
+
"loss": 0.0030770519375801085,
|
| 804 |
"step": 11000
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 7.660455486542443,
|
| 808 |
+
"grad_norm": 0.032615549862384796,
|
| 809 |
"learning_rate": 7.75257464659676e-05,
|
| 810 |
+
"loss": 0.0037489014863967894,
|
| 811 |
"step": 11100
|
| 812 |
},
|
| 813 |
{
|
| 814 |
"epoch": 7.729468599033816,
|
| 815 |
+
"grad_norm": 0.0029172301292419434,
|
| 816 |
"learning_rate": 7.702098581257692e-05,
|
| 817 |
+
"loss": 0.0034059333801269532,
|
| 818 |
"step": 11200
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 7.79848171152519,
|
| 822 |
+
"grad_norm": 0.02034895122051239,
|
| 823 |
"learning_rate": 7.651230490792762e-05,
|
| 824 |
+
"loss": 0.001802024245262146,
|
| 825 |
"step": 11300
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 7.867494824016563,
|
| 829 |
+
"grad_norm": 0.0322832353413105,
|
| 830 |
"learning_rate": 7.599977755232328e-05,
|
| 831 |
+
"loss": 0.0018968936800956727,
|
| 832 |
"step": 11400
|
| 833 |
},
|
| 834 |
{
|
| 835 |
"epoch": 7.936507936507937,
|
| 836 |
+
"grad_norm": 0.05175786465406418,
|
| 837 |
"learning_rate": 7.548347810411729e-05,
|
| 838 |
+
"loss": 0.002487382143735886,
|
| 839 |
"step": 11500
|
| 840 |
},
|
| 841 |
{
|
| 842 |
"epoch": 8.00552104899931,
|
| 843 |
+
"grad_norm": 0.0012496053241193295,
|
| 844 |
"learning_rate": 7.496348146892469e-05,
|
| 845 |
+
"loss": 0.0019437670707702637,
|
| 846 |
"step": 11600
|
| 847 |
},
|
| 848 |
{
|
| 849 |
"epoch": 8.074534161490684,
|
| 850 |
+
"grad_norm": 0.0010164726991206408,
|
| 851 |
"learning_rate": 7.443986308875479e-05,
|
| 852 |
+
"loss": 0.0019105494022369385,
|
| 853 |
"step": 11700
|
| 854 |
},
|
| 855 |
{
|
| 856 |
"epoch": 8.143547273982056,
|
| 857 |
+
"grad_norm": 0.00012283546675462276,
|
| 858 |
"learning_rate": 7.391269893106592e-05,
|
| 859 |
+
"loss": 0.001659536361694336,
|
| 860 |
"step": 11800
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 8.21256038647343,
|
| 864 |
+
"grad_norm": 0.03541138023138046,
|
| 865 |
"learning_rate": 7.338206547774395e-05,
|
| 866 |
+
"loss": 0.004322091042995453,
|
| 867 |
"step": 11900
|
| 868 |
},
|
| 869 |
{
|
| 870 |
"epoch": 8.281573498964804,
|
| 871 |
+
"grad_norm": 0.009761008433997631,
|
| 872 |
"learning_rate": 7.284803971400614e-05,
|
| 873 |
+
"loss": 0.0025467997789382934,
|
| 874 |
"step": 12000
|
| 875 |
},
|
| 876 |
{
|
| 877 |
"epoch": 8.281573498964804,
|
| 878 |
+
"eval_loss": 0.031825874000787735,
|
| 879 |
+
"eval_runtime": 239.4637,
|
| 880 |
+
"eval_samples_per_second": 23.507,
|
| 881 |
+
"eval_steps_per_second": 5.88,
|
| 882 |
"step": 12000
|
| 883 |
}
|
| 884 |
],
|
RuleMaze/quest/checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5560
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:029719a95557edd9724ff361e0cbd56cde434c96af79f5832c4af95872a253d4
|
| 3 |
size 5560
|