Instructions to use FluffyAIcode/Kakeya-OProver-Stage2-SFT1000 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use FluffyAIcode/Kakeya-OProver-Stage2-SFT1000 with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("m-a-p/OProver-8B") model = PeftModel.from_pretrained(base_model, "FluffyAIcode/Kakeya-OProver-Stage2-SFT1000") - Notebooks
- Google Colab
- Kaggle
Publish sanitized Stage 2 SFT1000 release
Browse files- .gitattributes +1 -0
- LICENSE +201 -0
- NOTICE +3 -0
- README.md +83 -0
- adapter_config.json +50 -0
- adapter_model.safetensors +3 -0
- chat_template.jinja +89 -0
- evaluation/results.json +19 -0
- provenance/manifest.json +26 -0
- tokenizer.json +3 -0
- tokenizer_config.json +15 -0
- training/metrics.jsonl +92 -0
- training/trainer_state.json +963 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
LICENSE
ADDED
|
@@ -0,0 +1,201 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Apache License
|
| 2 |
+
Version 2.0, January 2004
|
| 3 |
+
http://www.apache.org/licenses/
|
| 4 |
+
|
| 5 |
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
| 6 |
+
|
| 7 |
+
1. Definitions.
|
| 8 |
+
|
| 9 |
+
"License" shall mean the terms and conditions for use, reproduction,
|
| 10 |
+
and distribution as defined by Sections 1 through 9 of this document.
|
| 11 |
+
|
| 12 |
+
"Licensor" shall mean the copyright owner or entity authorized by
|
| 13 |
+
the copyright owner that is granting the License.
|
| 14 |
+
|
| 15 |
+
"Legal Entity" shall mean the union of the acting entity and all
|
| 16 |
+
other entities that control, are controlled by, or are under common
|
| 17 |
+
control with that entity. For the purposes of this definition,
|
| 18 |
+
"control" means (i) the power, direct or indirect, to cause the
|
| 19 |
+
direction or management of such entity, whether by contract or
|
| 20 |
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
| 21 |
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
| 22 |
+
|
| 23 |
+
"You" (or "Your") shall mean an individual or Legal Entity
|
| 24 |
+
exercising permissions granted by this License.
|
| 25 |
+
|
| 26 |
+
"Source" form shall mean the preferred form for making modifications,
|
| 27 |
+
including but not limited to software source code, documentation
|
| 28 |
+
source, and configuration files.
|
| 29 |
+
|
| 30 |
+
"Object" form shall mean any form resulting from mechanical
|
| 31 |
+
transformation or translation of a Source form, including but
|
| 32 |
+
not limited to compiled object code, generated documentation,
|
| 33 |
+
and conversions to other media types.
|
| 34 |
+
|
| 35 |
+
"Work" shall mean the work of authorship, whether in Source or
|
| 36 |
+
Object form, made available under the License, as indicated by a
|
| 37 |
+
copyright notice that is included in or attached to the work
|
| 38 |
+
(an example is provided in the Appendix below).
|
| 39 |
+
|
| 40 |
+
"Derivative Works" shall mean any work, whether in Source or Object
|
| 41 |
+
form, that is based on (or derived from) the Work and for which the
|
| 42 |
+
editorial revisions, annotations, elaborations, or other modifications
|
| 43 |
+
represent, as a whole, an original work of authorship. For the purposes
|
| 44 |
+
of this License, Derivative Works shall not include works that remain
|
| 45 |
+
separable from, or merely link (or bind by name) to the interfaces of,
|
| 46 |
+
the Work and Derivative Works thereof.
|
| 47 |
+
|
| 48 |
+
"Contribution" shall mean any work of authorship, including
|
| 49 |
+
the original version of the Work and any modifications or additions
|
| 50 |
+
to that Work or Derivative Works thereof, that is intentionally
|
| 51 |
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
| 52 |
+
or by an individual or Legal Entity authorized to submit on behalf of
|
| 53 |
+
the copyright owner. For the purposes of this definition, "submitted"
|
| 54 |
+
means any form of electronic, verbal, or written communication sent
|
| 55 |
+
to the Licensor or its representatives, including but not limited to
|
| 56 |
+
communication on electronic mailing lists, source code control systems,
|
| 57 |
+
and issue tracking systems that are managed by, or on behalf of, the
|
| 58 |
+
Licensor for the purpose of discussing and improving the Work, but
|
| 59 |
+
excluding communication that is conspicuously marked or otherwise
|
| 60 |
+
designated in writing by the copyright owner as "Not a Contribution."
|
| 61 |
+
|
| 62 |
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
| 63 |
+
on behalf of whom a Contribution has been received by Licensor and
|
| 64 |
+
subsequently incorporated within the Work.
|
| 65 |
+
|
| 66 |
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
| 67 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 68 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 69 |
+
copyright license to reproduce, prepare Derivative Works of,
|
| 70 |
+
publicly display, publicly perform, sublicense, and distribute the
|
| 71 |
+
Work and such Derivative Works in Source or Object form.
|
| 72 |
+
|
| 73 |
+
3. Grant of Patent License. Subject to the terms and conditions of
|
| 74 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 75 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 76 |
+
(except as stated in this section) patent license to make, have made,
|
| 77 |
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
| 78 |
+
where such license applies only to those patent claims licensable
|
| 79 |
+
by such Contributor that are necessarily infringed by their
|
| 80 |
+
Contribution(s) alone or by combination of their Contribution(s)
|
| 81 |
+
with the Work to which such Contribution(s) was submitted. If You
|
| 82 |
+
institute patent litigation against any entity (including a
|
| 83 |
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
| 84 |
+
or a Contribution incorporated within the Work constitutes direct
|
| 85 |
+
or contributory patent infringement, then any patent licenses
|
| 86 |
+
granted to You under this License for that Work shall terminate
|
| 87 |
+
as of the date such litigation is filed.
|
| 88 |
+
|
| 89 |
+
4. Redistribution. You may reproduce and distribute copies of the
|
| 90 |
+
Work or Derivative Works thereof in any medium, with or without
|
| 91 |
+
modifications, and in Source or Object form, provided that You
|
| 92 |
+
meet the following conditions:
|
| 93 |
+
|
| 94 |
+
(a) You must give any other recipients of the Work or
|
| 95 |
+
Derivative Works a copy of this License; and
|
| 96 |
+
|
| 97 |
+
(b) You must cause any modified files to carry prominent notices
|
| 98 |
+
stating that You changed the files; and
|
| 99 |
+
|
| 100 |
+
(c) You must retain, in the Source form of any Derivative Works
|
| 101 |
+
that You distribute, all copyright, patent, trademark, and
|
| 102 |
+
attribution notices from the Source form of the Work,
|
| 103 |
+
excluding those notices that do not pertain to any part of
|
| 104 |
+
the Derivative Works; and
|
| 105 |
+
|
| 106 |
+
(d) If the Work includes a "NOTICE" text file as part of its
|
| 107 |
+
distribution, then any Derivative Works that You distribute must
|
| 108 |
+
include a readable copy of the attribution notices contained
|
| 109 |
+
within such NOTICE file, excluding those notices that do not
|
| 110 |
+
pertain to any part of the Derivative Works, in at least one
|
| 111 |
+
of the following places: within a NOTICE text file distributed
|
| 112 |
+
as part of the Derivative Works; within the Source form or
|
| 113 |
+
documentation, if provided along with the Derivative Works; or,
|
| 114 |
+
within a display generated by the Derivative Works, if and
|
| 115 |
+
wherever such third-party notices normally appear. The contents
|
| 116 |
+
of the NOTICE file are for informational purposes only and
|
| 117 |
+
do not modify the License. You may add Your own attribution
|
| 118 |
+
notices within Derivative Works that You distribute, alongside
|
| 119 |
+
or as an addendum to the NOTICE text from the Work, provided
|
| 120 |
+
that such additional attribution notices cannot be construed
|
| 121 |
+
as modifying the License.
|
| 122 |
+
|
| 123 |
+
You may add Your own copyright statement to Your modifications and
|
| 124 |
+
may provide additional or different license terms and conditions
|
| 125 |
+
for use, reproduction, or distribution of Your modifications, or
|
| 126 |
+
for any such Derivative Works as a whole, provided Your use,
|
| 127 |
+
reproduction, and distribution of the Work otherwise complies with
|
| 128 |
+
the conditions stated in this License.
|
| 129 |
+
|
| 130 |
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
| 131 |
+
any Contribution intentionally submitted for inclusion in the Work
|
| 132 |
+
by You to the Licensor shall be under the terms and conditions of
|
| 133 |
+
this License, without any additional terms or conditions.
|
| 134 |
+
Notwithstanding the above, nothing herein shall supersede or modify
|
| 135 |
+
the terms of any separate license agreement you may have executed
|
| 136 |
+
with Licensor regarding such Contributions.
|
| 137 |
+
|
| 138 |
+
6. Trademarks. This License does not grant permission to use the trade
|
| 139 |
+
names, trademarks, service marks, or product names of the Licensor,
|
| 140 |
+
except as required for reasonable and customary use in describing the
|
| 141 |
+
origin of the Work and reproducing the content of the NOTICE file.
|
| 142 |
+
|
| 143 |
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
| 144 |
+
agreed to in writing, Licensor provides the Work (and each
|
| 145 |
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
| 146 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
| 147 |
+
implied, including, without limitation, any warranties or conditions
|
| 148 |
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
| 149 |
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
| 150 |
+
appropriateness of using or redistributing the Work and assume any
|
| 151 |
+
risks associated with Your exercise of permissions under this License.
|
| 152 |
+
|
| 153 |
+
8. Limitation of Liability. In no event and under no legal theory,
|
| 154 |
+
whether in tort (including negligence), contract, or otherwise,
|
| 155 |
+
unless required by applicable law (such as deliberate and grossly
|
| 156 |
+
negligent acts) or agreed to in writing, shall any Contributor be
|
| 157 |
+
liable to You for damages, including any direct, indirect, special,
|
| 158 |
+
incidental, or consequential damages of any character arising as a
|
| 159 |
+
result of this License or out of the use or inability to use the
|
| 160 |
+
Work (including but not limited to damages for loss of goodwill,
|
| 161 |
+
work stoppage, computer failure or malfunction, or any and all
|
| 162 |
+
other commercial damages or losses), even if such Contributor
|
| 163 |
+
has been advised of the possibility of such damages.
|
| 164 |
+
|
| 165 |
+
9. Accepting Warranty or Additional Liability. While redistributing
|
| 166 |
+
the Work or Derivative Works thereof, You may choose to offer,
|
| 167 |
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
| 168 |
+
or other liability obligations and/or rights consistent with this
|
| 169 |
+
License. However, in accepting such obligations, You may act only
|
| 170 |
+
on Your own behalf and on Your sole responsibility, not on behalf
|
| 171 |
+
of any other Contributor, and only if You agree to indemnify,
|
| 172 |
+
defend, and hold each Contributor harmless for any liability
|
| 173 |
+
incurred by, or claims asserted against, such Contributor by reason
|
| 174 |
+
of your accepting any such warranty or additional liability.
|
| 175 |
+
|
| 176 |
+
END OF TERMS AND CONDITIONS
|
| 177 |
+
|
| 178 |
+
APPENDIX: How to apply the Apache License to your work.
|
| 179 |
+
|
| 180 |
+
To apply the Apache License to your work, attach the following
|
| 181 |
+
boilerplate notice, with the fields enclosed by brackets "{}"
|
| 182 |
+
replaced with your own identifying information. (Don't include
|
| 183 |
+
the brackets!) The text should be enclosed in the appropriate
|
| 184 |
+
comment syntax for the file format. We also recommend that a
|
| 185 |
+
file or class name and description of purpose be included on the
|
| 186 |
+
same "printed page" as the copyright notice for easier
|
| 187 |
+
identification within third-party archives.
|
| 188 |
+
|
| 189 |
+
Copyright {yyyy} {name of copyright owner}
|
| 190 |
+
|
| 191 |
+
Licensed under the Apache License, Version 2.0 (the "License");
|
| 192 |
+
you may not use this file except in compliance with the License.
|
| 193 |
+
You may obtain a copy of the License at
|
| 194 |
+
|
| 195 |
+
http://www.apache.org/licenses/LICENSE-2.0
|
| 196 |
+
|
| 197 |
+
Unless required by applicable law or agreed to in writing, software
|
| 198 |
+
distributed under the License is distributed on an "AS IS" BASIS,
|
| 199 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
| 200 |
+
See the License for the specific language governing permissions and
|
| 201 |
+
limitations under the License.
|
NOTICE
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Kakeya OProver LoRA derivative, 2026 FluffyAIcode.
|
| 2 |
+
Base model: m-a-p/OProver-8B, Apache-2.0, revision cd9ffd383b584d95bf00e04b88b35b05928b211c.
|
| 3 |
+
Training material derives from leanprover-community/mathlib4, Apache-2.0, revision 360da6fa66c1273b76b6b2d8c5666fd5ac2e3b56.
|
README.md
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: m-a-p/OProver-8B
|
| 4 |
+
library_name: peft
|
| 5 |
+
pipeline_tag: text-generation
|
| 6 |
+
tags:
|
| 7 |
+
- lean
|
| 8 |
+
- lean4
|
| 9 |
+
- theorem-proving
|
| 10 |
+
- lora
|
| 11 |
+
- peft
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# Kakeya OProver Stage 2 — SFT1000
|
| 15 |
+
|
| 16 |
+
> **Status: MATCHED BASELINE**
|
| 17 |
+
|
| 18 |
+
## Summary
|
| 19 |
+
|
| 20 |
+
This is an independent PEFT LoRA checkpoint in the Kakeya OProver training journey. It is a proof-advisor research artifact, not an autonomous proof or mathematical truth claim.
|
| 21 |
+
|
| 22 |
+
## Training Data
|
| 23 |
+
|
| 24 |
+
Public data config: [`sft-1000`](https://huggingface.co/datasets/FluffyAIcode/Kakeya-OProver-Training). Artifact SHA-256: `5ebec3f7ed71d0af202924ea9d5ca12a857c5363b89b028dbf74268bf2e32f24`. Only Mathlib Apache-2.0 material and explicitly derived training negatives are included; provenance-incomplete OProofs mixtures were excluded.
|
| 25 |
+
|
| 26 |
+
## Method
|
| 27 |
+
|
| 28 |
+
QLoRA supervised fine-tuning on 700 Mathlib examples, with 100 validation and 200 excluded holdout examples.
|
| 29 |
+
|
| 30 |
+
## Strict Lean Evaluation
|
| 31 |
+
|
| 32 |
+
All matched Stage2–Stage4 evaluations used 22 tasks (18 holdout + 4 canary), one frozen task order, seed, temperature, token budget, and context policy. Parser rate was 22/22 and unsafe generations were 0 for matched evaluations.
|
| 33 |
+
|
| 34 |
+
```json
|
| 35 |
+
{
|
| 36 |
+
"status": "MATCHED BASELINE",
|
| 37 |
+
"all22": {
|
| 38 |
+
"exact": 9,
|
| 39 |
+
"n": 22,
|
| 40 |
+
"parser": 22,
|
| 41 |
+
"unsafe": 0
|
| 42 |
+
},
|
| 43 |
+
"holdout18": {
|
| 44 |
+
"exact": 5,
|
| 45 |
+
"n": 18
|
| 46 |
+
},
|
| 47 |
+
"canary4": {
|
| 48 |
+
"exact": 4,
|
| 49 |
+
"n": 4
|
| 50 |
+
},
|
| 51 |
+
"audit_sha256": "5e0f757353ffe17315c9b78c60bb46de9ae34e736f9460e69ec26ca1d6213f9c",
|
| 52 |
+
"protocol_root": "6047aa9e4f3ff9594f7913e943cf40dc050b49328b7a42a80a72beefe5c641d7"
|
| 53 |
+
}
|
| 54 |
+
```
|
| 55 |
+
|
| 56 |
+
## Safety
|
| 57 |
+
|
| 58 |
+
No private chain-of-thought, production prompts, Jensen candidate proof bodies, credentials, or raw candidate-bearing signed audits are published. Stage status constrains use; Stage4 is **PROOF-ADVISOR-ONLY**.
|
| 59 |
+
|
| 60 |
+
## Known Failures
|
| 61 |
+
|
| 62 |
+
5/18 holdout tasks failed exact Lean validation. Warning-bearing outputs and one length stop were observed; unsafe outputs were zero.
|
| 63 |
+
|
| 64 |
+
## Intended Use
|
| 65 |
+
|
| 66 |
+
Research on Lean proof suggestion, evaluation, and repair. Every generated proof must be compiled and independently reviewed. Not intended for unsupervised theorem claims or production deployment without separate authorization.
|
| 67 |
+
|
| 68 |
+
## Reproducibility
|
| 69 |
+
|
| 70 |
+
- Base: [`m-a-p/OProver-8B`](https://huggingface.co/m-a-p/OProver-8B) at `cd9ffd383b584d95bf00e04b88b35b05928b211c`
|
| 71 |
+
- Immutable source revision: `8eead931f4b090bca2421fead7d0c64191bea116`
|
| 72 |
+
- Adapter SHA-256: `5e5399518c0741140a14d2cca2fade8ec3eb95fb587139c5c9894d5e297a3b55`
|
| 73 |
+
- Dataset artifact SHA-256: `5ebec3f7ed71d0af202924ea9d5ca12a857c5363b89b028dbf74268bf2e32f24`
|
| 74 |
+
- Sanitized machine-readable files: `evaluation/results.json`, `training/metrics.jsonl`, `training/trainer_state.json`, `provenance/manifest.json`
|
| 75 |
+
- Evaluation/audit digest references are in `evaluation/results.json`; raw candidate-bearing artifacts remain private.
|
| 76 |
+
|
| 77 |
+
## Limitations
|
| 78 |
+
|
| 79 |
+
Small, single-seed strict suites cannot establish broad theorem-proving capability. Exact Lean rates have wide confidence intervals. Stage4 did not significantly outperform Stage2 in the matched paired test. Jensen remained `SEARCH_EXHAUSTED` with 0/6 Lean-valid candidates. **RH PROVED: NO.**
|
| 80 |
+
|
| 81 |
+
## Attribution
|
| 82 |
+
|
| 83 |
+
Base model: m-a-p/OProver-8B (Apache-2.0). Training data derives from Mathlib at `360da6fa66c1273b76b6b2d8c5666fd5ac2e3b56` (Apache-2.0). See `NOTICE`.
|
adapter_config.json
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alora_invocation_tokens": null,
|
| 3 |
+
"alpha_pattern": {},
|
| 4 |
+
"arrow_config": null,
|
| 5 |
+
"auto_mapping": null,
|
| 6 |
+
"base_model_name_or_path": "m-a-p/OProver-8B",
|
| 7 |
+
"bias": "none",
|
| 8 |
+
"corda_config": null,
|
| 9 |
+
"ensure_weight_tying": false,
|
| 10 |
+
"eva_config": null,
|
| 11 |
+
"exclude_modules": null,
|
| 12 |
+
"fan_in_fan_out": false,
|
| 13 |
+
"inference_mode": true,
|
| 14 |
+
"init_lora_weights": true,
|
| 15 |
+
"layer_replication": null,
|
| 16 |
+
"layers_pattern": null,
|
| 17 |
+
"layers_to_transform": null,
|
| 18 |
+
"loftq_config": {},
|
| 19 |
+
"lora_alpha": 64,
|
| 20 |
+
"lora_bias": false,
|
| 21 |
+
"lora_dropout": 0.05,
|
| 22 |
+
"lora_ga_config": null,
|
| 23 |
+
"megatron_config": null,
|
| 24 |
+
"megatron_core": "megatron.core",
|
| 25 |
+
"modules_to_save": null,
|
| 26 |
+
"monteclora_config": null,
|
| 27 |
+
"peft_type": "LORA",
|
| 28 |
+
"peft_version": "0.20.0",
|
| 29 |
+
"qalora_group_size": 16,
|
| 30 |
+
"r": 32,
|
| 31 |
+
"rank_pattern": {},
|
| 32 |
+
"revision": null,
|
| 33 |
+
"target_modules": [
|
| 34 |
+
"o_proj",
|
| 35 |
+
"k_proj",
|
| 36 |
+
"gate_proj",
|
| 37 |
+
"up_proj",
|
| 38 |
+
"q_proj",
|
| 39 |
+
"v_proj",
|
| 40 |
+
"down_proj"
|
| 41 |
+
],
|
| 42 |
+
"target_parameters": null,
|
| 43 |
+
"task_type": "CAUSAL_LM",
|
| 44 |
+
"trainable_token_indices": null,
|
| 45 |
+
"use_bdlora": null,
|
| 46 |
+
"use_dora": false,
|
| 47 |
+
"use_qalora": false,
|
| 48 |
+
"use_rslora": false,
|
| 49 |
+
"velora_config": null
|
| 50 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e5399518c0741140a14d2cca2fade8ec3eb95fb587139c5c9894d5e297a3b55
|
| 3 |
+
size 174656040
|
chat_template.jinja
ADDED
|
@@ -0,0 +1,89 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0].role == 'system' %}
|
| 4 |
+
{{- messages[0].content + '\n\n' }}
|
| 5 |
+
{%- endif %}
|
| 6 |
+
{{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 7 |
+
{%- for tool in tools %}
|
| 8 |
+
{{- "\n" }}
|
| 9 |
+
{{- tool | tojson }}
|
| 10 |
+
{%- endfor %}
|
| 11 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 12 |
+
{%- else %}
|
| 13 |
+
{%- if messages[0].role == 'system' %}
|
| 14 |
+
{{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
|
| 15 |
+
{%- endif %}
|
| 16 |
+
{%- endif %}
|
| 17 |
+
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
|
| 18 |
+
{%- for message in messages[::-1] %}
|
| 19 |
+
{%- set index = (messages|length - 1) - loop.index0 %}
|
| 20 |
+
{%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
|
| 21 |
+
{%- set ns.multi_step_tool = false %}
|
| 22 |
+
{%- set ns.last_query_index = index %}
|
| 23 |
+
{%- endif %}
|
| 24 |
+
{%- endfor %}
|
| 25 |
+
{%- for message in messages %}
|
| 26 |
+
{%- if message.content is string %}
|
| 27 |
+
{%- set content = message.content %}
|
| 28 |
+
{%- else %}
|
| 29 |
+
{%- set content = '' %}
|
| 30 |
+
{%- endif %}
|
| 31 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
|
| 32 |
+
{{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
|
| 33 |
+
{%- elif message.role == "assistant" %}
|
| 34 |
+
{%- set reasoning_content = '' %}
|
| 35 |
+
{%- if message.reasoning_content is string %}
|
| 36 |
+
{%- set reasoning_content = message.reasoning_content %}
|
| 37 |
+
{%- else %}
|
| 38 |
+
{%- if '</think>' in content %}
|
| 39 |
+
{%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
|
| 40 |
+
{%- set content = content.split('</think>')[-1].lstrip('\n') %}
|
| 41 |
+
{%- endif %}
|
| 42 |
+
{%- endif %}
|
| 43 |
+
{%- if loop.index0 > ns.last_query_index %}
|
| 44 |
+
{%- if loop.last or (not loop.last and reasoning_content) %}
|
| 45 |
+
{{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
|
| 46 |
+
{%- else %}
|
| 47 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 48 |
+
{%- endif %}
|
| 49 |
+
{%- else %}
|
| 50 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 51 |
+
{%- endif %}
|
| 52 |
+
{%- if message.tool_calls %}
|
| 53 |
+
{%- for tool_call in message.tool_calls %}
|
| 54 |
+
{%- if (loop.first and content) or (not loop.first) %}
|
| 55 |
+
{{- '\n' }}
|
| 56 |
+
{%- endif %}
|
| 57 |
+
{%- if tool_call.function %}
|
| 58 |
+
{%- set tool_call = tool_call.function %}
|
| 59 |
+
{%- endif %}
|
| 60 |
+
{{- '<tool_call>\n{"name": "' }}
|
| 61 |
+
{{- tool_call.name }}
|
| 62 |
+
{{- '", "arguments": ' }}
|
| 63 |
+
{%- if tool_call.arguments is string %}
|
| 64 |
+
{{- tool_call.arguments }}
|
| 65 |
+
{%- else %}
|
| 66 |
+
{{- tool_call.arguments | tojson }}
|
| 67 |
+
{%- endif %}
|
| 68 |
+
{{- '}\n</tool_call>' }}
|
| 69 |
+
{%- endfor %}
|
| 70 |
+
{%- endif %}
|
| 71 |
+
{{- '<|im_end|>\n' }}
|
| 72 |
+
{%- elif message.role == "tool" %}
|
| 73 |
+
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
|
| 74 |
+
{{- '<|im_start|>user' }}
|
| 75 |
+
{%- endif %}
|
| 76 |
+
{{- '\n<tool_response>\n' }}
|
| 77 |
+
{{- content }}
|
| 78 |
+
{{- '\n</tool_response>' }}
|
| 79 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 80 |
+
{{- '<|im_end|>\n' }}
|
| 81 |
+
{%- endif %}
|
| 82 |
+
{%- endif %}
|
| 83 |
+
{%- endfor %}
|
| 84 |
+
{%- if add_generation_prompt %}
|
| 85 |
+
{{- '<|im_start|>assistant\n' }}
|
| 86 |
+
{%- if enable_thinking is defined and enable_thinking is false %}
|
| 87 |
+
{{- '<think>\n\n</think>\n\n' }}
|
| 88 |
+
{%- endif %}
|
| 89 |
+
{%- endif %}
|
evaluation/results.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"status": "MATCHED BASELINE",
|
| 3 |
+
"all22": {
|
| 4 |
+
"exact": 9,
|
| 5 |
+
"n": 22,
|
| 6 |
+
"parser": 22,
|
| 7 |
+
"unsafe": 0
|
| 8 |
+
},
|
| 9 |
+
"holdout18": {
|
| 10 |
+
"exact": 5,
|
| 11 |
+
"n": 18
|
| 12 |
+
},
|
| 13 |
+
"canary4": {
|
| 14 |
+
"exact": 4,
|
| 15 |
+
"n": 4
|
| 16 |
+
},
|
| 17 |
+
"audit_sha256": "5e0f757353ffe17315c9b78c60bb46de9ae34e736f9460e69ec26ca1d6213f9c",
|
| 18 |
+
"protocol_root": "6047aa9e4f3ff9594f7913e943cf40dc050b49328b7a42a80a72beefe5c641d7"
|
| 19 |
+
}
|
provenance/manifest.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"schema_version": 1,
|
| 3 |
+
"public_repo": "FluffyAIcode/Kakeya-OProver-Stage2-SFT1000",
|
| 4 |
+
"base_model": "m-a-p/OProver-8B",
|
| 5 |
+
"base_revision": "cd9ffd383b584d95bf00e04b88b35b05928b211c",
|
| 6 |
+
"source_private_repo": "FluffyAIcode/oprover-8b-mathlib-lora-stage1",
|
| 7 |
+
"source_immutable_revision": "8eead931f4b090bca2421fead7d0c64191bea116",
|
| 8 |
+
"adapter_sha256": "5e5399518c0741140a14d2cca2fade8ec3eb95fb587139c5c9894d5e297a3b55",
|
| 9 |
+
"dataset_artifact_sha256": "5ebec3f7ed71d0af202924ea9d5ca12a857c5363b89b028dbf74268bf2e32f24",
|
| 10 |
+
"public_private_boundary": {
|
| 11 |
+
"public": [
|
| 12 |
+
"adapter",
|
| 13 |
+
"tokenizer",
|
| 14 |
+
"sanitized metrics/trainer state",
|
| 15 |
+
"aggregate strict evaluation",
|
| 16 |
+
"hashes/signatures/config"
|
| 17 |
+
],
|
| 18 |
+
"private": [
|
| 19 |
+
"chain-of-thought",
|
| 20 |
+
"production prompts",
|
| 21 |
+
"Jensen candidate proof bodies",
|
| 22 |
+
"raw candidate-bearing signed audits",
|
| 23 |
+
"optimizer and RNG checkpoints"
|
| 24 |
+
]
|
| 25 |
+
}
|
| 26 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ab5a866c748c81fe21eb29faa5c87c5f6448415348e5ab984ba2d69c6d68b36
|
| 3 |
+
size 11428194
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": null,
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|im_end|>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"model_max_length": 131072,
|
| 11 |
+
"pad_token": "<|endoftext|>",
|
| 12 |
+
"split_special_tokens": false,
|
| 13 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 14 |
+
"unk_token": null
|
| 15 |
+
}
|
training/metrics.jsonl
ADDED
|
@@ -0,0 +1,92 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"entropy": 0.571594450622797, "epoch": 0.022857142857142857, "grad_norm": 2.03125, "learning_rate": 0.0, "loss": 1.892868161201477, "mean_token_accuracy": 0.6494692210108042, "num_tokens": 28665.0, "step": 1}
|
| 2 |
+
{"entropy": 0.5400022976100445, "epoch": 0.045714285714285714, "grad_norm": 1.953125, "learning_rate": 5e-05, "loss": 1.3431090116500854, "mean_token_accuracy": 0.6841957792639732, "num_tokens": 56298.0, "step": 2}
|
| 3 |
+
{"entropy": 0.5547708570957184, "epoch": 0.06857142857142857, "grad_norm": 2.046875, "learning_rate": 0.0001, "loss": 1.7731074094772339, "mean_token_accuracy": 0.6730335894972086, "num_tokens": 87134.0, "step": 3}
|
| 4 |
+
{"entropy": 0.6138171833008528, "epoch": 0.09142857142857143, "grad_norm": 1.890625, "learning_rate": 0.00015000000000000001, "loss": 1.6849099397659302, "mean_token_accuracy": 0.7163603082299232, "num_tokens": 110491.0, "step": 4}
|
| 5 |
+
{"entropy": 0.713952723890543, "epoch": 0.11428571428571428, "grad_norm": 1.609375, "learning_rate": 0.0002, "loss": 1.6369731426239014, "mean_token_accuracy": 0.6762979347258806, "num_tokens": 138858.0, "step": 5}
|
| 6 |
+
{"entropy": 0.9489956572651863, "epoch": 0.13714285714285715, "grad_norm": 1.609375, "learning_rate": 0.00019993007047883988, "loss": 1.6043992042541504, "mean_token_accuracy": 0.6542574372142553, "num_tokens": 164666.0, "step": 6}
|
| 7 |
+
{"entropy": 1.2052904553711414, "epoch": 0.16, "grad_norm": 2.046875, "learning_rate": 0.00019972037971811802, "loss": 1.1030031442642212, "mean_token_accuracy": 0.7115810438990593, "num_tokens": 198168.0, "step": 7}
|
| 8 |
+
{"entropy": 1.5162354111671448, "epoch": 0.18285714285714286, "grad_norm": 2.609375, "learning_rate": 0.00019937122098932428, "loss": 1.338626503944397, "mean_token_accuracy": 0.7028692364692688, "num_tokens": 229142.0, "step": 8}
|
| 9 |
+
{"entropy": 1.6831717118620872, "epoch": 0.2057142857142857, "grad_norm": 1.203125, "learning_rate": 0.00019888308262251285, "loss": 1.5491583347320557, "mean_token_accuracy": 0.669173551723361, "num_tokens": 257334.0, "step": 9}
|
| 10 |
+
{"entropy": 1.2227612249553204, "epoch": 0.22857142857142856, "grad_norm": 0.6328125, "learning_rate": 0.00019825664732332884, "loss": 1.0137903690338135, "mean_token_accuracy": 0.7204974591732025, "num_tokens": 286904.0, "step": 10}
|
| 11 |
+
{"entropy": 1.1638432145118713, "epoch": 0.25142857142857145, "grad_norm": 0.83984375, "learning_rate": 0.00019749279121818235, "loss": 1.0442439317703247, "mean_token_accuracy": 0.7225705273449421, "num_tokens": 313566.0, "step": 11}
|
| 12 |
+
{"entropy": 1.090791016817093, "epoch": 0.2742857142857143, "grad_norm": 0.75, "learning_rate": 0.00019659258262890683, "loss": 1.2202626466751099, "mean_token_accuracy": 0.6869607474654913, "num_tokens": 343912.0, "step": 12}
|
| 13 |
+
{"entropy": 1.100559327751398, "epoch": 0.29714285714285715, "grad_norm": 0.73046875, "learning_rate": 0.0001955572805786141, "loss": 1.4237300157546997, "mean_token_accuracy": 0.7026770114898682, "num_tokens": 367824.0, "step": 13}
|
| 14 |
+
{"entropy": 1.0772063918411732, "epoch": 0.32, "grad_norm": 0.734375, "learning_rate": 0.00019438833303083678, "loss": 1.3612886667251587, "mean_token_accuracy": 0.7167600486427546, "num_tokens": 397629.0, "step": 14}
|
| 15 |
+
{"entropy": 1.0418964736163616, "epoch": 0.34285714285714286, "grad_norm": 0.76171875, "learning_rate": 0.00019308737486442045, "loss": 1.2377914190292358, "mean_token_accuracy": 0.7337369881570339, "num_tokens": 426172.0, "step": 15}
|
| 16 |
+
{"entropy": 1.02627420052886, "epoch": 0.3657142857142857, "grad_norm": 0.640625, "learning_rate": 0.00019165622558699763, "loss": 1.043224811553955, "mean_token_accuracy": 0.757489874958992, "num_tokens": 455884.0, "step": 16}
|
| 17 |
+
{"entropy": 1.0985115952789783, "epoch": 0.38857142857142857, "grad_norm": 0.75390625, "learning_rate": 0.0001900968867902419, "loss": 1.0292891263961792, "mean_token_accuracy": 0.7834405265748501, "num_tokens": 479943.0, "step": 17}
|
| 18 |
+
{"entropy": 0.9926935620605946, "epoch": 0.4114285714285714, "grad_norm": 0.5234375, "learning_rate": 0.00018841153935046098, "loss": 1.0027269124984741, "mean_token_accuracy": 0.7445859499275684, "num_tokens": 512401.0, "step": 18}
|
| 19 |
+
{"entropy": 1.184921432286501, "epoch": 0.4342857142857143, "grad_norm": 0.6953125, "learning_rate": 0.00018660254037844388, "loss": 0.8860166072845459, "mean_token_accuracy": 0.7711772620677948, "num_tokens": 532314.0, "step": 19}
|
| 20 |
+
{"entropy": 1.0456415861845016, "epoch": 0.45714285714285713, "grad_norm": 0.578125, "learning_rate": 0.00018467241992282843, "loss": 0.821132242679596, "mean_token_accuracy": 0.7591653987765312, "num_tokens": 560611.0, "step": 20}
|
| 21 |
+
{"entropy": 1.0742540583014488, "epoch": 0.48, "grad_norm": 0.6640625, "learning_rate": 0.0001826238774315995, "loss": 1.1304452419281006, "mean_token_accuracy": 0.7381897754967213, "num_tokens": 582239.0, "step": 21}
|
| 22 |
+
{"entropy": 1.0069492012262344, "epoch": 0.5028571428571429, "grad_norm": 0.83203125, "learning_rate": 0.00018045977797666684, "loss": 1.0569063425064087, "mean_token_accuracy": 0.7475782930850983, "num_tokens": 605785.0, "step": 22}
|
| 23 |
+
{"epoch": 0.5028571428571429, "eval_entropy": 1.0555952689051629, "eval_loss": 1.0447686910629272, "eval_mean_token_accuracy": 0.75280682772398, "eval_num_tokens": 605785.0, "eval_runtime": 77.2364, "eval_samples_per_second": 1.295, "eval_steps_per_second": 1.295, "step": 22}
|
| 24 |
+
{"entropy": 1.0209653116762638, "epoch": 0.5257142857142857, "grad_norm": 0.6953125, "learning_rate": 0.000178183148246803, "loss": 0.9215176105499268, "mean_token_accuracy": 0.7393636666238308, "num_tokens": 630668.0, "step": 23}
|
| 25 |
+
{"entropy": 1.1484412215650082, "epoch": 0.5485714285714286, "grad_norm": 0.63671875, "learning_rate": 0.0001757971723145453, "loss": 1.0585880279541016, "mean_token_accuracy": 0.6772188358008862, "num_tokens": 651568.0, "step": 24}
|
| 26 |
+
{"entropy": 1.022590946406126, "epoch": 0.5714285714285714, "grad_norm": 0.6640625, "learning_rate": 0.00017330518718298264, "loss": 0.8019014596939087, "mean_token_accuracy": 0.7844364829361439, "num_tokens": 677010.0, "step": 25}
|
| 27 |
+
{"entropy": 1.1501350328326225, "epoch": 0.5942857142857143, "grad_norm": 0.578125, "learning_rate": 0.00017071067811865476, "loss": 1.1763538122177124, "mean_token_accuracy": 0.7165477126836777, "num_tokens": 699559.0, "step": 26}
|
| 28 |
+
{"entropy": 1.1520305536687374, "epoch": 0.6171428571428571, "grad_norm": 0.84765625, "learning_rate": 0.00016801727377709194, "loss": 1.3568904399871826, "mean_token_accuracy": 0.7278081495314837, "num_tokens": 723860.0, "step": 27}
|
| 29 |
+
{"entropy": 1.1409958936274052, "epoch": 0.64, "grad_norm": 0.73828125, "learning_rate": 0.00016522874112781213, "loss": 1.2788770198822021, "mean_token_accuracy": 0.688488058745861, "num_tokens": 743869.0, "step": 28}
|
| 30 |
+
{"entropy": 1.06721768155694, "epoch": 0.6628571428571428, "grad_norm": 0.796875, "learning_rate": 0.00016234898018587337, "loss": 0.9878795742988586, "mean_token_accuracy": 0.7843082323670387, "num_tokens": 767668.0, "step": 29}
|
| 31 |
+
{"entropy": 0.9738106857985258, "epoch": 0.6857142857142857, "grad_norm": 0.578125, "learning_rate": 0.00015938201855735014, "loss": 0.7424539923667908, "mean_token_accuracy": 0.7907082177698612, "num_tokens": 796481.0, "step": 30}
|
| 32 |
+
{"entropy": 0.9771843180060387, "epoch": 0.7085714285714285, "grad_norm": 0.75, "learning_rate": 0.0001563320058063622, "loss": 0.9626051187515259, "mean_token_accuracy": 0.7523193173110485, "num_tokens": 829764.0, "step": 31}
|
| 33 |
+
{"entropy": 1.151510078459978, "epoch": 0.7314285714285714, "grad_norm": 0.78125, "learning_rate": 0.00015320320765153367, "loss": 1.270058512687683, "mean_token_accuracy": 0.6807492543011904, "num_tokens": 855562.0, "step": 32}
|
| 34 |
+
{"entropy": 0.9595269281417131, "epoch": 0.7542857142857143, "grad_norm": 0.5625, "learning_rate": 0.00015000000000000001, "loss": 0.9262505173683167, "mean_token_accuracy": 0.7448003850877285, "num_tokens": 886448.0, "step": 33}
|
| 35 |
+
{"entropy": 0.9837835170328617, "epoch": 0.7771428571428571, "grad_norm": 0.82421875, "learning_rate": 0.0001467268628273062, "loss": 0.7896201014518738, "mean_token_accuracy": 0.7692355290055275, "num_tokens": 909179.0, "step": 34}
|
| 36 |
+
{"entropy": 1.0101780369877815, "epoch": 0.8, "grad_norm": 0.9375, "learning_rate": 0.00014338837391175582, "loss": 1.1079224348068237, "mean_token_accuracy": 0.743992704898119, "num_tokens": 932293.0, "step": 35}
|
| 37 |
+
{"entropy": 1.124088928103447, "epoch": 0.8228571428571428, "grad_norm": 0.84375, "learning_rate": 0.00013998920243197407, "loss": 1.2773886919021606, "mean_token_accuracy": 0.7306922785937786, "num_tokens": 953785.0, "step": 36}
|
| 38 |
+
{"entropy": 1.0273678377270699, "epoch": 0.8457142857142858, "grad_norm": 0.765625, "learning_rate": 0.00013653410243663952, "loss": 1.0014641284942627, "mean_token_accuracy": 0.7343494817614555, "num_tokens": 983541.0, "step": 37}
|
| 39 |
+
{"entropy": 1.1427535712718964, "epoch": 0.8685714285714285, "grad_norm": 0.671875, "learning_rate": 0.00013302790619551674, "loss": 1.1199904680252075, "mean_token_accuracy": 0.7464530095458031, "num_tokens": 1009431.0, "step": 38}
|
| 40 |
+
{"entropy": 1.1668125055730343, "epoch": 0.8914285714285715, "grad_norm": 0.6015625, "learning_rate": 0.00012947551744109043, "loss": 0.9645350575447083, "mean_token_accuracy": 0.780300073325634, "num_tokens": 1027701.0, "step": 39}
|
| 41 |
+
{"entropy": 1.1839856766164303, "epoch": 0.9142857142857143, "grad_norm": 0.62109375, "learning_rate": 0.00012588190451025207, "loss": 1.2035199403762817, "mean_token_accuracy": 0.7434030883014202, "num_tokens": 1055642.0, "step": 40}
|
| 42 |
+
{"entropy": 1.103015311062336, "epoch": 0.9371428571428572, "grad_norm": 0.62109375, "learning_rate": 0.00012225209339563145, "loss": 1.0058553218841553, "mean_token_accuracy": 0.7543482556939125, "num_tokens": 1084745.0, "step": 41}
|
| 43 |
+
{"entropy": 1.1135993152856827, "epoch": 0.96, "grad_norm": 0.6953125, "learning_rate": 0.00011859116071629149, "loss": 1.2151912450790405, "mean_token_accuracy": 0.7335421219468117, "num_tokens": 1110320.0, "step": 42}
|
| 44 |
+
{"entropy": 1.1436597369611263, "epoch": 0.9828571428571429, "grad_norm": 0.58203125, "learning_rate": 0.00011490422661761744, "loss": 1.0186842679977417, "mean_token_accuracy": 0.7390037253499031, "num_tokens": 1134964.0, "step": 43}
|
| 45 |
+
{"entropy": 1.078730583190918, "epoch": 1.0, "grad_norm": 0.65625, "learning_rate": 0.00011119644761033078, "loss": 0.8101186752319336, "mean_token_accuracy": 0.7597745011250178, "num_tokens": 1153583.0, "step": 44}
|
| 46 |
+
{"epoch": 1.0, "eval_entropy": 1.0479114320874214, "eval_loss": 0.9729008674621582, "eval_mean_token_accuracy": 0.7565665817260743, "eval_num_tokens": 1153583.0, "eval_runtime": 77.2335, "eval_samples_per_second": 1.295, "eval_steps_per_second": 1.295, "step": 44}
|
| 47 |
+
{"entropy": 1.0368036814033985, "epoch": 1.022857142857143, "grad_norm": 0.56640625, "learning_rate": 0.00010747300935864243, "loss": 0.8468512296676636, "mean_token_accuracy": 0.8111352436244488, "num_tokens": 1174609.0, "step": 45}
|
| 48 |
+
{"entropy": 0.9628574140369892, "epoch": 1.0457142857142858, "grad_norm": 0.5546875, "learning_rate": 0.0001037391194276326, "loss": 0.5208293199539185, "mean_token_accuracy": 0.8187313564121723, "num_tokens": 1201897.0, "step": 46}
|
| 49 |
+
{"entropy": 1.0299121290445328, "epoch": 1.0685714285714285, "grad_norm": 0.765625, "learning_rate": 0.0001, "loss": 0.7912545204162598, "mean_token_accuracy": 0.7928437106311321, "num_tokens": 1230046.0, "step": 47}
|
| 50 |
+
{"entropy": 1.0342555530369282, "epoch": 1.0914285714285714, "grad_norm": 0.74609375, "learning_rate": 9.626088057236745e-05, "loss": 0.9557701349258423, "mean_token_accuracy": 0.7545025758445263, "num_tokens": 1257406.0, "step": 48}
|
| 51 |
+
{"entropy": 0.9161418825387955, "epoch": 1.1142857142857143, "grad_norm": 0.7109375, "learning_rate": 9.252699064135758e-05, "loss": 1.0390901565551758, "mean_token_accuracy": 0.7715597599744797, "num_tokens": 1291055.0, "step": 49}
|
| 52 |
+
{"entropy": 0.8812923058867455, "epoch": 1.1371428571428572, "grad_norm": 0.625, "learning_rate": 8.880355238966923e-05, "loss": 0.6828033328056335, "mean_token_accuracy": 0.8140444047749043, "num_tokens": 1318257.0, "step": 50}
|
| 53 |
+
{"entropy": 1.0039622746407986, "epoch": 1.16, "grad_norm": 0.546875, "learning_rate": 8.509577338238255e-05, "loss": 0.7247803807258606, "mean_token_accuracy": 0.7645618040114641, "num_tokens": 1341106.0, "step": 51}
|
| 54 |
+
{"entropy": 0.905966442078352, "epoch": 1.1828571428571428, "grad_norm": 0.5625, "learning_rate": 8.140883928370855e-05, "loss": 0.7403977513313293, "mean_token_accuracy": 0.8016074895858765, "num_tokens": 1369760.0, "step": 52}
|
| 55 |
+
{"entropy": 0.9338806178420782, "epoch": 1.2057142857142857, "grad_norm": 0.490234375, "learning_rate": 7.774790660436858e-05, "loss": 0.6406869888305664, "mean_token_accuracy": 0.7956290915608406, "num_tokens": 1392567.0, "step": 53}
|
| 56 |
+
{"entropy": 0.8846337422728539, "epoch": 1.2285714285714286, "grad_norm": 0.69921875, "learning_rate": 7.411809548974792e-05, "loss": 0.8057032823562622, "mean_token_accuracy": 0.7973924726247787, "num_tokens": 1419324.0, "step": 54}
|
| 57 |
+
{"entropy": 0.9206812232732773, "epoch": 1.2514285714285713, "grad_norm": 0.8125, "learning_rate": 7.052448255890957e-05, "loss": 0.6728847026824951, "mean_token_accuracy": 0.8473470285534859, "num_tokens": 1446756.0, "step": 55}
|
| 58 |
+
{"entropy": 0.9307001233100891, "epoch": 1.2742857142857142, "grad_norm": 0.8515625, "learning_rate": 6.697209380448333e-05, "loss": 0.7187720537185669, "mean_token_accuracy": 0.8393540307879448, "num_tokens": 1468062.0, "step": 56}
|
| 59 |
+
{"entropy": 0.8503624759614468, "epoch": 1.2971428571428572, "grad_norm": 0.64453125, "learning_rate": 6.34658975633605e-05, "loss": 0.6491850018501282, "mean_token_accuracy": 0.834420669823885, "num_tokens": 1497763.0, "step": 57}
|
| 60 |
+
{"entropy": 0.8430384919047356, "epoch": 1.32, "grad_norm": 0.62890625, "learning_rate": 6.001079756802592e-05, "loss": 0.7592802047729492, "mean_token_accuracy": 0.7834662459790707, "num_tokens": 1523103.0, "step": 58}
|
| 61 |
+
{"entropy": 1.0481715574860573, "epoch": 1.342857142857143, "grad_norm": 1.15625, "learning_rate": 5.6611626088244194e-05, "loss": 0.8737943768501282, "mean_token_accuracy": 0.7567208111286163, "num_tokens": 1540582.0, "step": 59}
|
| 62 |
+
{"entropy": 0.849721547216177, "epoch": 1.3657142857142857, "grad_norm": 0.8984375, "learning_rate": 5.32731371726938e-05, "loss": 0.7518897652626038, "mean_token_accuracy": 0.8164225146174431, "num_tokens": 1562096.0, "step": 60}
|
| 63 |
+
{"entropy": 0.7319265194237232, "epoch": 1.3885714285714286, "grad_norm": 0.76953125, "learning_rate": 5.000000000000002e-05, "loss": 0.5088726282119751, "mean_token_accuracy": 0.8368714712560177, "num_tokens": 1588869.0, "step": 61}
|
| 64 |
+
{"entropy": 0.7714136429131031, "epoch": 1.4114285714285715, "grad_norm": 0.515625, "learning_rate": 4.6796792348466356e-05, "loss": 0.5990605354309082, "mean_token_accuracy": 0.8109956867992878, "num_tokens": 1618753.0, "step": 62}
|
| 65 |
+
{"entropy": 0.8922704569995403, "epoch": 1.4342857142857142, "grad_norm": 0.7265625, "learning_rate": 4.3667994193637796e-05, "loss": 0.7471989393234253, "mean_token_accuracy": 0.8137485198676586, "num_tokens": 1639035.0, "step": 63}
|
| 66 |
+
{"entropy": 0.8474112637341022, "epoch": 1.457142857142857, "grad_norm": 0.9375, "learning_rate": 4.0617981442649855e-05, "loss": 0.9250304102897644, "mean_token_accuracy": 0.7889886423945427, "num_tokens": 1667695.0, "step": 64}
|
| 67 |
+
{"entropy": 0.8302664868533611, "epoch": 1.48, "grad_norm": 0.8984375, "learning_rate": 3.7651019814126654e-05, "loss": 0.9558159112930298, "mean_token_accuracy": 0.7763096019625664, "num_tokens": 1691022.0, "step": 65}
|
| 68 |
+
{"entropy": 0.8031172640621662, "epoch": 1.502857142857143, "grad_norm": 0.79296875, "learning_rate": 3.477125887218792e-05, "loss": 0.801313042640686, "mean_token_accuracy": 0.8177301362156868, "num_tokens": 1724435.0, "step": 66}
|
| 69 |
+
{"epoch": 1.502857142857143, "eval_entropy": 0.8457768812775612, "eval_loss": 0.9875430464744568, "eval_mean_token_accuracy": 0.756916128396988, "eval_num_tokens": 1724435.0, "eval_runtime": 77.2316, "eval_samples_per_second": 1.295, "eval_steps_per_second": 1.295, "step": 66}
|
| 70 |
+
{"entropy": 0.841597568243742, "epoch": 1.5257142857142858, "grad_norm": 0.8359375, "learning_rate": 3.198272622290804e-05, "loss": 0.8447511196136475, "mean_token_accuracy": 0.7982683703303337, "num_tokens": 1752489.0, "step": 67}
|
| 71 |
+
{"entropy": 0.873002614825964, "epoch": 1.5485714285714285, "grad_norm": 0.68359375, "learning_rate": 2.9289321881345254e-05, "loss": 0.8038796782493591, "mean_token_accuracy": 0.8067350275814533, "num_tokens": 1784291.0, "step": 68}
|
| 72 |
+
{"entropy": 0.9177370071411133, "epoch": 1.5714285714285714, "grad_norm": 0.6796875, "learning_rate": 2.669481281701739e-05, "loss": 0.8557254672050476, "mean_token_accuracy": 0.8137136548757553, "num_tokens": 1804179.0, "step": 69}
|
| 73 |
+
{"entropy": 0.8276135660707951, "epoch": 1.5942857142857143, "grad_norm": 0.73828125, "learning_rate": 2.420282768545469e-05, "loss": 0.6800186634063721, "mean_token_accuracy": 0.8346021547913551, "num_tokens": 1831612.0, "step": 70}
|
| 74 |
+
{"entropy": 0.9364416636526585, "epoch": 1.617142857142857, "grad_norm": 0.75390625, "learning_rate": 2.181685175319702e-05, "loss": 0.7834956049919128, "mean_token_accuracy": 0.7997728437185287, "num_tokens": 1853595.0, "step": 71}
|
| 75 |
+
{"entropy": 0.8710121884942055, "epoch": 1.6400000000000001, "grad_norm": 0.765625, "learning_rate": 1.9540222023333166e-05, "loss": 0.6057661771774292, "mean_token_accuracy": 0.8205794207751751, "num_tokens": 1882408.0, "step": 72}
|
| 76 |
+
{"entropy": 0.8379442691802979, "epoch": 1.6628571428571428, "grad_norm": 0.7734375, "learning_rate": 1.7376122568400532e-05, "loss": 0.7106757760047913, "mean_token_accuracy": 0.8180459216237068, "num_tokens": 1904944.0, "step": 73}
|
| 77 |
+
{"entropy": 0.8745108097791672, "epoch": 1.6857142857142857, "grad_norm": 1.2109375, "learning_rate": 1.5327580077171587e-05, "loss": 0.9369223713874817, "mean_token_accuracy": 0.8044208958745003, "num_tokens": 1933402.0, "step": 74}
|
| 78 |
+
{"entropy": 0.8298410475254059, "epoch": 1.7085714285714286, "grad_norm": 0.8828125, "learning_rate": 1.339745962155613e-05, "loss": 0.6346684694290161, "mean_token_accuracy": 0.8335375674068928, "num_tokens": 1960093.0, "step": 75}
|
| 79 |
+
{"entropy": 0.8175692446529865, "epoch": 1.7314285714285713, "grad_norm": 0.91796875, "learning_rate": 1.1588460649539035e-05, "loss": 0.9119294285774231, "mean_token_accuracy": 0.8193050436675549, "num_tokens": 1986339.0, "step": 76}
|
| 80 |
+
{"entropy": 0.8975704163312912, "epoch": 1.7542857142857144, "grad_norm": 0.859375, "learning_rate": 9.903113209758096e-06, "loss": 0.7237526178359985, "mean_token_accuracy": 0.7882252000272274, "num_tokens": 2011176.0, "step": 77}
|
| 81 |
+
{"entropy": 0.8363670147955418, "epoch": 1.7771428571428571, "grad_norm": 0.74609375, "learning_rate": 8.343774413002381e-06, "loss": 0.6269868016242981, "mean_token_accuracy": 0.8103100657463074, "num_tokens": 2041289.0, "step": 78}
|
| 82 |
+
{"entropy": 0.8557109721004963, "epoch": 1.8, "grad_norm": 0.6171875, "learning_rate": 6.9126251355795864e-06, "loss": 0.5580326914787292, "mean_token_accuracy": 0.8539069853723049, "num_tokens": 2069418.0, "step": 79}
|
| 83 |
+
{"entropy": 0.790475644171238, "epoch": 1.822857142857143, "grad_norm": 0.984375, "learning_rate": 5.611666969163243e-06, "loss": 0.6902757287025452, "mean_token_accuracy": 0.8313734903931618, "num_tokens": 2101162.0, "step": 80}
|
| 84 |
+
{"entropy": 0.8487820513546467, "epoch": 1.8457142857142856, "grad_norm": 0.7265625, "learning_rate": 4.442719421385922e-06, "loss": 0.6248161792755127, "mean_token_accuracy": 0.8196811378002167, "num_tokens": 2129262.0, "step": 81}
|
| 85 |
+
{"entropy": 0.8948404788970947, "epoch": 1.8685714285714285, "grad_norm": 0.87890625, "learning_rate": 3.40741737109318e-06, "loss": 0.8848119378089905, "mean_token_accuracy": 0.7821721211075783, "num_tokens": 2152597.0, "step": 82}
|
| 86 |
+
{"entropy": 0.9239097908139229, "epoch": 1.8914285714285715, "grad_norm": 0.7421875, "learning_rate": 2.5072087818176382e-06, "loss": 0.7065685987472534, "mean_token_accuracy": 0.807807132601738, "num_tokens": 2177161.0, "step": 83}
|
| 87 |
+
{"entropy": 0.7839450426399708, "epoch": 1.9142857142857141, "grad_norm": 0.77734375, "learning_rate": 1.7433526766711728e-06, "loss": 0.7359437942504883, "mean_token_accuracy": 0.8403091952204704, "num_tokens": 2209858.0, "step": 84}
|
| 88 |
+
{"entropy": 0.843034666031599, "epoch": 1.9371428571428573, "grad_norm": 0.765625, "learning_rate": 1.1169173774871478e-06, "loss": 0.7045976519584656, "mean_token_accuracy": 0.7968335002660751, "num_tokens": 2238762.0, "step": 85}
|
| 89 |
+
{"entropy": 0.8172502107918262, "epoch": 1.96, "grad_norm": 0.73046875, "learning_rate": 6.287790106757396e-07, "loss": 0.7153176665306091, "mean_token_accuracy": 0.8268864750862122, "num_tokens": 2266000.0, "step": 86}
|
| 90 |
+
{"entropy": 0.9393381252884865, "epoch": 1.9828571428571429, "grad_norm": 0.8046875, "learning_rate": 2.7962028188198706e-07, "loss": 0.7222880721092224, "mean_token_accuracy": 0.8043639399111271, "num_tokens": 2291896.0, "step": 87}
|
| 91 |
+
{"entropy": 0.9007064700126648, "epoch": 2.0, "grad_norm": 0.9453125, "learning_rate": 6.992952116013918e-08, "loss": 0.9412174224853516, "mean_token_accuracy": 0.7673612783352534, "num_tokens": 2307166.0, "step": 88}
|
| 92 |
+
{"epoch": 2.0, "eval_entropy": 0.8574838742613793, "eval_loss": 0.9815174341201782, "eval_mean_token_accuracy": 0.7545835164189338, "eval_num_tokens": 2307166.0, "eval_runtime": 77.2359, "eval_samples_per_second": 1.295, "eval_steps_per_second": 1.295, "step": 88}
|
training/trainer_state.json
ADDED
|
@@ -0,0 +1,963 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"max_steps": 88,
|
| 3 |
+
"save_steps": 22,
|
| 4 |
+
"is_world_process_zero": true,
|
| 5 |
+
"train_batch_size": 1,
|
| 6 |
+
"stateful_callbacks": {
|
| 7 |
+
"TrainerControl": {
|
| 8 |
+
"args": {
|
| 9 |
+
"should_epoch_stop": false,
|
| 10 |
+
"should_evaluate": false,
|
| 11 |
+
"should_log": false,
|
| 12 |
+
"should_save": true,
|
| 13 |
+
"should_training_stop": true
|
| 14 |
+
},
|
| 15 |
+
"attributes": {}
|
| 16 |
+
}
|
| 17 |
+
},
|
| 18 |
+
"is_hyper_param_search": false,
|
| 19 |
+
"num_input_tokens_seen": 0,
|
| 20 |
+
"log_history": [
|
| 21 |
+
{
|
| 22 |
+
"entropy": 0.571594450622797,
|
| 23 |
+
"epoch": 0.022857142857142857,
|
| 24 |
+
"grad_norm": 2.03125,
|
| 25 |
+
"learning_rate": 0.0,
|
| 26 |
+
"loss": 1.892868161201477,
|
| 27 |
+
"mean_token_accuracy": 0.6494692210108042,
|
| 28 |
+
"num_tokens": 28665.0,
|
| 29 |
+
"step": 1
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"entropy": 0.5400022976100445,
|
| 33 |
+
"epoch": 0.045714285714285714,
|
| 34 |
+
"grad_norm": 1.953125,
|
| 35 |
+
"learning_rate": 5e-05,
|
| 36 |
+
"loss": 1.3431090116500854,
|
| 37 |
+
"mean_token_accuracy": 0.6841957792639732,
|
| 38 |
+
"num_tokens": 56298.0,
|
| 39 |
+
"step": 2
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"entropy": 0.5547708570957184,
|
| 43 |
+
"epoch": 0.06857142857142857,
|
| 44 |
+
"grad_norm": 2.046875,
|
| 45 |
+
"learning_rate": 0.0001,
|
| 46 |
+
"loss": 1.7731074094772339,
|
| 47 |
+
"mean_token_accuracy": 0.6730335894972086,
|
| 48 |
+
"num_tokens": 87134.0,
|
| 49 |
+
"step": 3
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"entropy": 0.6138171833008528,
|
| 53 |
+
"epoch": 0.09142857142857143,
|
| 54 |
+
"grad_norm": 1.890625,
|
| 55 |
+
"learning_rate": 0.00015000000000000001,
|
| 56 |
+
"loss": 1.6849099397659302,
|
| 57 |
+
"mean_token_accuracy": 0.7163603082299232,
|
| 58 |
+
"num_tokens": 110491.0,
|
| 59 |
+
"step": 4
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"entropy": 0.713952723890543,
|
| 63 |
+
"epoch": 0.11428571428571428,
|
| 64 |
+
"grad_norm": 1.609375,
|
| 65 |
+
"learning_rate": 0.0002,
|
| 66 |
+
"loss": 1.6369731426239014,
|
| 67 |
+
"mean_token_accuracy": 0.6762979347258806,
|
| 68 |
+
"num_tokens": 138858.0,
|
| 69 |
+
"step": 5
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"entropy": 0.9489956572651863,
|
| 73 |
+
"epoch": 0.13714285714285715,
|
| 74 |
+
"grad_norm": 1.609375,
|
| 75 |
+
"learning_rate": 0.00019993007047883988,
|
| 76 |
+
"loss": 1.6043992042541504,
|
| 77 |
+
"mean_token_accuracy": 0.6542574372142553,
|
| 78 |
+
"num_tokens": 164666.0,
|
| 79 |
+
"step": 6
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"entropy": 1.2052904553711414,
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"grad_norm": 2.046875,
|
| 85 |
+
"learning_rate": 0.00019972037971811802,
|
| 86 |
+
"loss": 1.1030031442642212,
|
| 87 |
+
"mean_token_accuracy": 0.7115810438990593,
|
| 88 |
+
"num_tokens": 198168.0,
|
| 89 |
+
"step": 7
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"entropy": 1.5162354111671448,
|
| 93 |
+
"epoch": 0.18285714285714286,
|
| 94 |
+
"grad_norm": 2.609375,
|
| 95 |
+
"learning_rate": 0.00019937122098932428,
|
| 96 |
+
"loss": 1.338626503944397,
|
| 97 |
+
"mean_token_accuracy": 0.7028692364692688,
|
| 98 |
+
"num_tokens": 229142.0,
|
| 99 |
+
"step": 8
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"entropy": 1.6831717118620872,
|
| 103 |
+
"epoch": 0.2057142857142857,
|
| 104 |
+
"grad_norm": 1.203125,
|
| 105 |
+
"learning_rate": 0.00019888308262251285,
|
| 106 |
+
"loss": 1.5491583347320557,
|
| 107 |
+
"mean_token_accuracy": 0.669173551723361,
|
| 108 |
+
"num_tokens": 257334.0,
|
| 109 |
+
"step": 9
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"entropy": 1.2227612249553204,
|
| 113 |
+
"epoch": 0.22857142857142856,
|
| 114 |
+
"grad_norm": 0.6328125,
|
| 115 |
+
"learning_rate": 0.00019825664732332884,
|
| 116 |
+
"loss": 1.0137903690338135,
|
| 117 |
+
"mean_token_accuracy": 0.7204974591732025,
|
| 118 |
+
"num_tokens": 286904.0,
|
| 119 |
+
"step": 10
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"entropy": 1.1638432145118713,
|
| 123 |
+
"epoch": 0.25142857142857145,
|
| 124 |
+
"grad_norm": 0.83984375,
|
| 125 |
+
"learning_rate": 0.00019749279121818235,
|
| 126 |
+
"loss": 1.0442439317703247,
|
| 127 |
+
"mean_token_accuracy": 0.7225705273449421,
|
| 128 |
+
"num_tokens": 313566.0,
|
| 129 |
+
"step": 11
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"entropy": 1.090791016817093,
|
| 133 |
+
"epoch": 0.2742857142857143,
|
| 134 |
+
"grad_norm": 0.75,
|
| 135 |
+
"learning_rate": 0.00019659258262890683,
|
| 136 |
+
"loss": 1.2202626466751099,
|
| 137 |
+
"mean_token_accuracy": 0.6869607474654913,
|
| 138 |
+
"num_tokens": 343912.0,
|
| 139 |
+
"step": 12
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"entropy": 1.100559327751398,
|
| 143 |
+
"epoch": 0.29714285714285715,
|
| 144 |
+
"grad_norm": 0.73046875,
|
| 145 |
+
"learning_rate": 0.0001955572805786141,
|
| 146 |
+
"loss": 1.4237300157546997,
|
| 147 |
+
"mean_token_accuracy": 0.7026770114898682,
|
| 148 |
+
"num_tokens": 367824.0,
|
| 149 |
+
"step": 13
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"entropy": 1.0772063918411732,
|
| 153 |
+
"epoch": 0.32,
|
| 154 |
+
"grad_norm": 0.734375,
|
| 155 |
+
"learning_rate": 0.00019438833303083678,
|
| 156 |
+
"loss": 1.3612886667251587,
|
| 157 |
+
"mean_token_accuracy": 0.7167600486427546,
|
| 158 |
+
"num_tokens": 397629.0,
|
| 159 |
+
"step": 14
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"entropy": 1.0418964736163616,
|
| 163 |
+
"epoch": 0.34285714285714286,
|
| 164 |
+
"grad_norm": 0.76171875,
|
| 165 |
+
"learning_rate": 0.00019308737486442045,
|
| 166 |
+
"loss": 1.2377914190292358,
|
| 167 |
+
"mean_token_accuracy": 0.7337369881570339,
|
| 168 |
+
"num_tokens": 426172.0,
|
| 169 |
+
"step": 15
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"entropy": 1.02627420052886,
|
| 173 |
+
"epoch": 0.3657142857142857,
|
| 174 |
+
"grad_norm": 0.640625,
|
| 175 |
+
"learning_rate": 0.00019165622558699763,
|
| 176 |
+
"loss": 1.043224811553955,
|
| 177 |
+
"mean_token_accuracy": 0.757489874958992,
|
| 178 |
+
"num_tokens": 455884.0,
|
| 179 |
+
"step": 16
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"entropy": 1.0985115952789783,
|
| 183 |
+
"epoch": 0.38857142857142857,
|
| 184 |
+
"grad_norm": 0.75390625,
|
| 185 |
+
"learning_rate": 0.0001900968867902419,
|
| 186 |
+
"loss": 1.0292891263961792,
|
| 187 |
+
"mean_token_accuracy": 0.7834405265748501,
|
| 188 |
+
"num_tokens": 479943.0,
|
| 189 |
+
"step": 17
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"entropy": 0.9926935620605946,
|
| 193 |
+
"epoch": 0.4114285714285714,
|
| 194 |
+
"grad_norm": 0.5234375,
|
| 195 |
+
"learning_rate": 0.00018841153935046098,
|
| 196 |
+
"loss": 1.0027269124984741,
|
| 197 |
+
"mean_token_accuracy": 0.7445859499275684,
|
| 198 |
+
"num_tokens": 512401.0,
|
| 199 |
+
"step": 18
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"entropy": 1.184921432286501,
|
| 203 |
+
"epoch": 0.4342857142857143,
|
| 204 |
+
"grad_norm": 0.6953125,
|
| 205 |
+
"learning_rate": 0.00018660254037844388,
|
| 206 |
+
"loss": 0.8860166072845459,
|
| 207 |
+
"mean_token_accuracy": 0.7711772620677948,
|
| 208 |
+
"num_tokens": 532314.0,
|
| 209 |
+
"step": 19
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"entropy": 1.0456415861845016,
|
| 213 |
+
"epoch": 0.45714285714285713,
|
| 214 |
+
"grad_norm": 0.578125,
|
| 215 |
+
"learning_rate": 0.00018467241992282843,
|
| 216 |
+
"loss": 0.821132242679596,
|
| 217 |
+
"mean_token_accuracy": 0.7591653987765312,
|
| 218 |
+
"num_tokens": 560611.0,
|
| 219 |
+
"step": 20
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"entropy": 1.0742540583014488,
|
| 223 |
+
"epoch": 0.48,
|
| 224 |
+
"grad_norm": 0.6640625,
|
| 225 |
+
"learning_rate": 0.0001826238774315995,
|
| 226 |
+
"loss": 1.1304452419281006,
|
| 227 |
+
"mean_token_accuracy": 0.7381897754967213,
|
| 228 |
+
"num_tokens": 582239.0,
|
| 229 |
+
"step": 21
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"entropy": 1.0069492012262344,
|
| 233 |
+
"epoch": 0.5028571428571429,
|
| 234 |
+
"grad_norm": 0.83203125,
|
| 235 |
+
"learning_rate": 0.00018045977797666684,
|
| 236 |
+
"loss": 1.0569063425064087,
|
| 237 |
+
"mean_token_accuracy": 0.7475782930850983,
|
| 238 |
+
"num_tokens": 605785.0,
|
| 239 |
+
"step": 22
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.5028571428571429,
|
| 243 |
+
"eval_entropy": 1.0555952689051629,
|
| 244 |
+
"eval_loss": 1.0447686910629272,
|
| 245 |
+
"eval_mean_token_accuracy": 0.75280682772398,
|
| 246 |
+
"eval_num_tokens": 605785.0,
|
| 247 |
+
"eval_runtime": 77.2364,
|
| 248 |
+
"eval_samples_per_second": 1.295,
|
| 249 |
+
"eval_steps_per_second": 1.295,
|
| 250 |
+
"step": 22
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"entropy": 1.0209653116762638,
|
| 254 |
+
"epoch": 0.5257142857142857,
|
| 255 |
+
"grad_norm": 0.6953125,
|
| 256 |
+
"learning_rate": 0.000178183148246803,
|
| 257 |
+
"loss": 0.9215176105499268,
|
| 258 |
+
"mean_token_accuracy": 0.7393636666238308,
|
| 259 |
+
"num_tokens": 630668.0,
|
| 260 |
+
"step": 23
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"entropy": 1.1484412215650082,
|
| 264 |
+
"epoch": 0.5485714285714286,
|
| 265 |
+
"grad_norm": 0.63671875,
|
| 266 |
+
"learning_rate": 0.0001757971723145453,
|
| 267 |
+
"loss": 1.0585880279541016,
|
| 268 |
+
"mean_token_accuracy": 0.6772188358008862,
|
| 269 |
+
"num_tokens": 651568.0,
|
| 270 |
+
"step": 24
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"entropy": 1.022590946406126,
|
| 274 |
+
"epoch": 0.5714285714285714,
|
| 275 |
+
"grad_norm": 0.6640625,
|
| 276 |
+
"learning_rate": 0.00017330518718298264,
|
| 277 |
+
"loss": 0.8019014596939087,
|
| 278 |
+
"mean_token_accuracy": 0.7844364829361439,
|
| 279 |
+
"num_tokens": 677010.0,
|
| 280 |
+
"step": 25
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"entropy": 1.1501350328326225,
|
| 284 |
+
"epoch": 0.5942857142857143,
|
| 285 |
+
"grad_norm": 0.578125,
|
| 286 |
+
"learning_rate": 0.00017071067811865476,
|
| 287 |
+
"loss": 1.1763538122177124,
|
| 288 |
+
"mean_token_accuracy": 0.7165477126836777,
|
| 289 |
+
"num_tokens": 699559.0,
|
| 290 |
+
"step": 26
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"entropy": 1.1520305536687374,
|
| 294 |
+
"epoch": 0.6171428571428571,
|
| 295 |
+
"grad_norm": 0.84765625,
|
| 296 |
+
"learning_rate": 0.00016801727377709194,
|
| 297 |
+
"loss": 1.3568904399871826,
|
| 298 |
+
"mean_token_accuracy": 0.7278081495314837,
|
| 299 |
+
"num_tokens": 723860.0,
|
| 300 |
+
"step": 27
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"entropy": 1.1409958936274052,
|
| 304 |
+
"epoch": 0.64,
|
| 305 |
+
"grad_norm": 0.73828125,
|
| 306 |
+
"learning_rate": 0.00016522874112781213,
|
| 307 |
+
"loss": 1.2788770198822021,
|
| 308 |
+
"mean_token_accuracy": 0.688488058745861,
|
| 309 |
+
"num_tokens": 743869.0,
|
| 310 |
+
"step": 28
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"entropy": 1.06721768155694,
|
| 314 |
+
"epoch": 0.6628571428571428,
|
| 315 |
+
"grad_norm": 0.796875,
|
| 316 |
+
"learning_rate": 0.00016234898018587337,
|
| 317 |
+
"loss": 0.9878795742988586,
|
| 318 |
+
"mean_token_accuracy": 0.7843082323670387,
|
| 319 |
+
"num_tokens": 767668.0,
|
| 320 |
+
"step": 29
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"entropy": 0.9738106857985258,
|
| 324 |
+
"epoch": 0.6857142857142857,
|
| 325 |
+
"grad_norm": 0.578125,
|
| 326 |
+
"learning_rate": 0.00015938201855735014,
|
| 327 |
+
"loss": 0.7424539923667908,
|
| 328 |
+
"mean_token_accuracy": 0.7907082177698612,
|
| 329 |
+
"num_tokens": 796481.0,
|
| 330 |
+
"step": 30
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"entropy": 0.9771843180060387,
|
| 334 |
+
"epoch": 0.7085714285714285,
|
| 335 |
+
"grad_norm": 0.75,
|
| 336 |
+
"learning_rate": 0.0001563320058063622,
|
| 337 |
+
"loss": 0.9626051187515259,
|
| 338 |
+
"mean_token_accuracy": 0.7523193173110485,
|
| 339 |
+
"num_tokens": 829764.0,
|
| 340 |
+
"step": 31
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"entropy": 1.151510078459978,
|
| 344 |
+
"epoch": 0.7314285714285714,
|
| 345 |
+
"grad_norm": 0.78125,
|
| 346 |
+
"learning_rate": 0.00015320320765153367,
|
| 347 |
+
"loss": 1.270058512687683,
|
| 348 |
+
"mean_token_accuracy": 0.6807492543011904,
|
| 349 |
+
"num_tokens": 855562.0,
|
| 350 |
+
"step": 32
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"entropy": 0.9595269281417131,
|
| 354 |
+
"epoch": 0.7542857142857143,
|
| 355 |
+
"grad_norm": 0.5625,
|
| 356 |
+
"learning_rate": 0.00015000000000000001,
|
| 357 |
+
"loss": 0.9262505173683167,
|
| 358 |
+
"mean_token_accuracy": 0.7448003850877285,
|
| 359 |
+
"num_tokens": 886448.0,
|
| 360 |
+
"step": 33
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"entropy": 0.9837835170328617,
|
| 364 |
+
"epoch": 0.7771428571428571,
|
| 365 |
+
"grad_norm": 0.82421875,
|
| 366 |
+
"learning_rate": 0.0001467268628273062,
|
| 367 |
+
"loss": 0.7896201014518738,
|
| 368 |
+
"mean_token_accuracy": 0.7692355290055275,
|
| 369 |
+
"num_tokens": 909179.0,
|
| 370 |
+
"step": 34
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"entropy": 1.0101780369877815,
|
| 374 |
+
"epoch": 0.8,
|
| 375 |
+
"grad_norm": 0.9375,
|
| 376 |
+
"learning_rate": 0.00014338837391175582,
|
| 377 |
+
"loss": 1.1079224348068237,
|
| 378 |
+
"mean_token_accuracy": 0.743992704898119,
|
| 379 |
+
"num_tokens": 932293.0,
|
| 380 |
+
"step": 35
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"entropy": 1.124088928103447,
|
| 384 |
+
"epoch": 0.8228571428571428,
|
| 385 |
+
"grad_norm": 0.84375,
|
| 386 |
+
"learning_rate": 0.00013998920243197407,
|
| 387 |
+
"loss": 1.2773886919021606,
|
| 388 |
+
"mean_token_accuracy": 0.7306922785937786,
|
| 389 |
+
"num_tokens": 953785.0,
|
| 390 |
+
"step": 36
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"entropy": 1.0273678377270699,
|
| 394 |
+
"epoch": 0.8457142857142858,
|
| 395 |
+
"grad_norm": 0.765625,
|
| 396 |
+
"learning_rate": 0.00013653410243663952,
|
| 397 |
+
"loss": 1.0014641284942627,
|
| 398 |
+
"mean_token_accuracy": 0.7343494817614555,
|
| 399 |
+
"num_tokens": 983541.0,
|
| 400 |
+
"step": 37
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"entropy": 1.1427535712718964,
|
| 404 |
+
"epoch": 0.8685714285714285,
|
| 405 |
+
"grad_norm": 0.671875,
|
| 406 |
+
"learning_rate": 0.00013302790619551674,
|
| 407 |
+
"loss": 1.1199904680252075,
|
| 408 |
+
"mean_token_accuracy": 0.7464530095458031,
|
| 409 |
+
"num_tokens": 1009431.0,
|
| 410 |
+
"step": 38
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"entropy": 1.1668125055730343,
|
| 414 |
+
"epoch": 0.8914285714285715,
|
| 415 |
+
"grad_norm": 0.6015625,
|
| 416 |
+
"learning_rate": 0.00012947551744109043,
|
| 417 |
+
"loss": 0.9645350575447083,
|
| 418 |
+
"mean_token_accuracy": 0.780300073325634,
|
| 419 |
+
"num_tokens": 1027701.0,
|
| 420 |
+
"step": 39
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"entropy": 1.1839856766164303,
|
| 424 |
+
"epoch": 0.9142857142857143,
|
| 425 |
+
"grad_norm": 0.62109375,
|
| 426 |
+
"learning_rate": 0.00012588190451025207,
|
| 427 |
+
"loss": 1.2035199403762817,
|
| 428 |
+
"mean_token_accuracy": 0.7434030883014202,
|
| 429 |
+
"num_tokens": 1055642.0,
|
| 430 |
+
"step": 40
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"entropy": 1.103015311062336,
|
| 434 |
+
"epoch": 0.9371428571428572,
|
| 435 |
+
"grad_norm": 0.62109375,
|
| 436 |
+
"learning_rate": 0.00012225209339563145,
|
| 437 |
+
"loss": 1.0058553218841553,
|
| 438 |
+
"mean_token_accuracy": 0.7543482556939125,
|
| 439 |
+
"num_tokens": 1084745.0,
|
| 440 |
+
"step": 41
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"entropy": 1.1135993152856827,
|
| 444 |
+
"epoch": 0.96,
|
| 445 |
+
"grad_norm": 0.6953125,
|
| 446 |
+
"learning_rate": 0.00011859116071629149,
|
| 447 |
+
"loss": 1.2151912450790405,
|
| 448 |
+
"mean_token_accuracy": 0.7335421219468117,
|
| 449 |
+
"num_tokens": 1110320.0,
|
| 450 |
+
"step": 42
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"entropy": 1.1436597369611263,
|
| 454 |
+
"epoch": 0.9828571428571429,
|
| 455 |
+
"grad_norm": 0.58203125,
|
| 456 |
+
"learning_rate": 0.00011490422661761744,
|
| 457 |
+
"loss": 1.0186842679977417,
|
| 458 |
+
"mean_token_accuracy": 0.7390037253499031,
|
| 459 |
+
"num_tokens": 1134964.0,
|
| 460 |
+
"step": 43
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"entropy": 1.078730583190918,
|
| 464 |
+
"epoch": 1.0,
|
| 465 |
+
"grad_norm": 0.65625,
|
| 466 |
+
"learning_rate": 0.00011119644761033078,
|
| 467 |
+
"loss": 0.8101186752319336,
|
| 468 |
+
"mean_token_accuracy": 0.7597745011250178,
|
| 469 |
+
"num_tokens": 1153583.0,
|
| 470 |
+
"step": 44
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.0,
|
| 474 |
+
"eval_entropy": 1.0479114320874214,
|
| 475 |
+
"eval_loss": 0.9729008674621582,
|
| 476 |
+
"eval_mean_token_accuracy": 0.7565665817260743,
|
| 477 |
+
"eval_num_tokens": 1153583.0,
|
| 478 |
+
"eval_runtime": 77.2335,
|
| 479 |
+
"eval_samples_per_second": 1.295,
|
| 480 |
+
"eval_steps_per_second": 1.295,
|
| 481 |
+
"step": 44
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"entropy": 1.0368036814033985,
|
| 485 |
+
"epoch": 1.022857142857143,
|
| 486 |
+
"grad_norm": 0.56640625,
|
| 487 |
+
"learning_rate": 0.00010747300935864243,
|
| 488 |
+
"loss": 0.8468512296676636,
|
| 489 |
+
"mean_token_accuracy": 0.8111352436244488,
|
| 490 |
+
"num_tokens": 1174609.0,
|
| 491 |
+
"step": 45
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"entropy": 0.9628574140369892,
|
| 495 |
+
"epoch": 1.0457142857142858,
|
| 496 |
+
"grad_norm": 0.5546875,
|
| 497 |
+
"learning_rate": 0.0001037391194276326,
|
| 498 |
+
"loss": 0.5208293199539185,
|
| 499 |
+
"mean_token_accuracy": 0.8187313564121723,
|
| 500 |
+
"num_tokens": 1201897.0,
|
| 501 |
+
"step": 46
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"entropy": 1.0299121290445328,
|
| 505 |
+
"epoch": 1.0685714285714285,
|
| 506 |
+
"grad_norm": 0.765625,
|
| 507 |
+
"learning_rate": 0.0001,
|
| 508 |
+
"loss": 0.7912545204162598,
|
| 509 |
+
"mean_token_accuracy": 0.7928437106311321,
|
| 510 |
+
"num_tokens": 1230046.0,
|
| 511 |
+
"step": 47
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"entropy": 1.0342555530369282,
|
| 515 |
+
"epoch": 1.0914285714285714,
|
| 516 |
+
"grad_norm": 0.74609375,
|
| 517 |
+
"learning_rate": 9.626088057236745e-05,
|
| 518 |
+
"loss": 0.9557701349258423,
|
| 519 |
+
"mean_token_accuracy": 0.7545025758445263,
|
| 520 |
+
"num_tokens": 1257406.0,
|
| 521 |
+
"step": 48
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"entropy": 0.9161418825387955,
|
| 525 |
+
"epoch": 1.1142857142857143,
|
| 526 |
+
"grad_norm": 0.7109375,
|
| 527 |
+
"learning_rate": 9.252699064135758e-05,
|
| 528 |
+
"loss": 1.0390901565551758,
|
| 529 |
+
"mean_token_accuracy": 0.7715597599744797,
|
| 530 |
+
"num_tokens": 1291055.0,
|
| 531 |
+
"step": 49
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"entropy": 0.8812923058867455,
|
| 535 |
+
"epoch": 1.1371428571428572,
|
| 536 |
+
"grad_norm": 0.625,
|
| 537 |
+
"learning_rate": 8.880355238966923e-05,
|
| 538 |
+
"loss": 0.6828033328056335,
|
| 539 |
+
"mean_token_accuracy": 0.8140444047749043,
|
| 540 |
+
"num_tokens": 1318257.0,
|
| 541 |
+
"step": 50
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"entropy": 1.0039622746407986,
|
| 545 |
+
"epoch": 1.16,
|
| 546 |
+
"grad_norm": 0.546875,
|
| 547 |
+
"learning_rate": 8.509577338238255e-05,
|
| 548 |
+
"loss": 0.7247803807258606,
|
| 549 |
+
"mean_token_accuracy": 0.7645618040114641,
|
| 550 |
+
"num_tokens": 1341106.0,
|
| 551 |
+
"step": 51
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"entropy": 0.905966442078352,
|
| 555 |
+
"epoch": 1.1828571428571428,
|
| 556 |
+
"grad_norm": 0.5625,
|
| 557 |
+
"learning_rate": 8.140883928370855e-05,
|
| 558 |
+
"loss": 0.7403977513313293,
|
| 559 |
+
"mean_token_accuracy": 0.8016074895858765,
|
| 560 |
+
"num_tokens": 1369760.0,
|
| 561 |
+
"step": 52
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"entropy": 0.9338806178420782,
|
| 565 |
+
"epoch": 1.2057142857142857,
|
| 566 |
+
"grad_norm": 0.490234375,
|
| 567 |
+
"learning_rate": 7.774790660436858e-05,
|
| 568 |
+
"loss": 0.6406869888305664,
|
| 569 |
+
"mean_token_accuracy": 0.7956290915608406,
|
| 570 |
+
"num_tokens": 1392567.0,
|
| 571 |
+
"step": 53
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"entropy": 0.8846337422728539,
|
| 575 |
+
"epoch": 1.2285714285714286,
|
| 576 |
+
"grad_norm": 0.69921875,
|
| 577 |
+
"learning_rate": 7.411809548974792e-05,
|
| 578 |
+
"loss": 0.8057032823562622,
|
| 579 |
+
"mean_token_accuracy": 0.7973924726247787,
|
| 580 |
+
"num_tokens": 1419324.0,
|
| 581 |
+
"step": 54
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"entropy": 0.9206812232732773,
|
| 585 |
+
"epoch": 1.2514285714285713,
|
| 586 |
+
"grad_norm": 0.8125,
|
| 587 |
+
"learning_rate": 7.052448255890957e-05,
|
| 588 |
+
"loss": 0.6728847026824951,
|
| 589 |
+
"mean_token_accuracy": 0.8473470285534859,
|
| 590 |
+
"num_tokens": 1446756.0,
|
| 591 |
+
"step": 55
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"entropy": 0.9307001233100891,
|
| 595 |
+
"epoch": 1.2742857142857142,
|
| 596 |
+
"grad_norm": 0.8515625,
|
| 597 |
+
"learning_rate": 6.697209380448333e-05,
|
| 598 |
+
"loss": 0.7187720537185669,
|
| 599 |
+
"mean_token_accuracy": 0.8393540307879448,
|
| 600 |
+
"num_tokens": 1468062.0,
|
| 601 |
+
"step": 56
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"entropy": 0.8503624759614468,
|
| 605 |
+
"epoch": 1.2971428571428572,
|
| 606 |
+
"grad_norm": 0.64453125,
|
| 607 |
+
"learning_rate": 6.34658975633605e-05,
|
| 608 |
+
"loss": 0.6491850018501282,
|
| 609 |
+
"mean_token_accuracy": 0.834420669823885,
|
| 610 |
+
"num_tokens": 1497763.0,
|
| 611 |
+
"step": 57
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"entropy": 0.8430384919047356,
|
| 615 |
+
"epoch": 1.32,
|
| 616 |
+
"grad_norm": 0.62890625,
|
| 617 |
+
"learning_rate": 6.001079756802592e-05,
|
| 618 |
+
"loss": 0.7592802047729492,
|
| 619 |
+
"mean_token_accuracy": 0.7834662459790707,
|
| 620 |
+
"num_tokens": 1523103.0,
|
| 621 |
+
"step": 58
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"entropy": 1.0481715574860573,
|
| 625 |
+
"epoch": 1.342857142857143,
|
| 626 |
+
"grad_norm": 1.15625,
|
| 627 |
+
"learning_rate": 5.6611626088244194e-05,
|
| 628 |
+
"loss": 0.8737943768501282,
|
| 629 |
+
"mean_token_accuracy": 0.7567208111286163,
|
| 630 |
+
"num_tokens": 1540582.0,
|
| 631 |
+
"step": 59
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"entropy": 0.849721547216177,
|
| 635 |
+
"epoch": 1.3657142857142857,
|
| 636 |
+
"grad_norm": 0.8984375,
|
| 637 |
+
"learning_rate": 5.32731371726938e-05,
|
| 638 |
+
"loss": 0.7518897652626038,
|
| 639 |
+
"mean_token_accuracy": 0.8164225146174431,
|
| 640 |
+
"num_tokens": 1562096.0,
|
| 641 |
+
"step": 60
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"entropy": 0.7319265194237232,
|
| 645 |
+
"epoch": 1.3885714285714286,
|
| 646 |
+
"grad_norm": 0.76953125,
|
| 647 |
+
"learning_rate": 5.000000000000002e-05,
|
| 648 |
+
"loss": 0.5088726282119751,
|
| 649 |
+
"mean_token_accuracy": 0.8368714712560177,
|
| 650 |
+
"num_tokens": 1588869.0,
|
| 651 |
+
"step": 61
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"entropy": 0.7714136429131031,
|
| 655 |
+
"epoch": 1.4114285714285715,
|
| 656 |
+
"grad_norm": 0.515625,
|
| 657 |
+
"learning_rate": 4.6796792348466356e-05,
|
| 658 |
+
"loss": 0.5990605354309082,
|
| 659 |
+
"mean_token_accuracy": 0.8109956867992878,
|
| 660 |
+
"num_tokens": 1618753.0,
|
| 661 |
+
"step": 62
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"entropy": 0.8922704569995403,
|
| 665 |
+
"epoch": 1.4342857142857142,
|
| 666 |
+
"grad_norm": 0.7265625,
|
| 667 |
+
"learning_rate": 4.3667994193637796e-05,
|
| 668 |
+
"loss": 0.7471989393234253,
|
| 669 |
+
"mean_token_accuracy": 0.8137485198676586,
|
| 670 |
+
"num_tokens": 1639035.0,
|
| 671 |
+
"step": 63
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"entropy": 0.8474112637341022,
|
| 675 |
+
"epoch": 1.457142857142857,
|
| 676 |
+
"grad_norm": 0.9375,
|
| 677 |
+
"learning_rate": 4.0617981442649855e-05,
|
| 678 |
+
"loss": 0.9250304102897644,
|
| 679 |
+
"mean_token_accuracy": 0.7889886423945427,
|
| 680 |
+
"num_tokens": 1667695.0,
|
| 681 |
+
"step": 64
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"entropy": 0.8302664868533611,
|
| 685 |
+
"epoch": 1.48,
|
| 686 |
+
"grad_norm": 0.8984375,
|
| 687 |
+
"learning_rate": 3.7651019814126654e-05,
|
| 688 |
+
"loss": 0.9558159112930298,
|
| 689 |
+
"mean_token_accuracy": 0.7763096019625664,
|
| 690 |
+
"num_tokens": 1691022.0,
|
| 691 |
+
"step": 65
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"entropy": 0.8031172640621662,
|
| 695 |
+
"epoch": 1.502857142857143,
|
| 696 |
+
"grad_norm": 0.79296875,
|
| 697 |
+
"learning_rate": 3.477125887218792e-05,
|
| 698 |
+
"loss": 0.801313042640686,
|
| 699 |
+
"mean_token_accuracy": 0.8177301362156868,
|
| 700 |
+
"num_tokens": 1724435.0,
|
| 701 |
+
"step": 66
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.502857142857143,
|
| 705 |
+
"eval_entropy": 0.8457768812775612,
|
| 706 |
+
"eval_loss": 0.9875430464744568,
|
| 707 |
+
"eval_mean_token_accuracy": 0.756916128396988,
|
| 708 |
+
"eval_num_tokens": 1724435.0,
|
| 709 |
+
"eval_runtime": 77.2316,
|
| 710 |
+
"eval_samples_per_second": 1.295,
|
| 711 |
+
"eval_steps_per_second": 1.295,
|
| 712 |
+
"step": 66
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"entropy": 0.841597568243742,
|
| 716 |
+
"epoch": 1.5257142857142858,
|
| 717 |
+
"grad_norm": 0.8359375,
|
| 718 |
+
"learning_rate": 3.198272622290804e-05,
|
| 719 |
+
"loss": 0.8447511196136475,
|
| 720 |
+
"mean_token_accuracy": 0.7982683703303337,
|
| 721 |
+
"num_tokens": 1752489.0,
|
| 722 |
+
"step": 67
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"entropy": 0.873002614825964,
|
| 726 |
+
"epoch": 1.5485714285714285,
|
| 727 |
+
"grad_norm": 0.68359375,
|
| 728 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 729 |
+
"loss": 0.8038796782493591,
|
| 730 |
+
"mean_token_accuracy": 0.8067350275814533,
|
| 731 |
+
"num_tokens": 1784291.0,
|
| 732 |
+
"step": 68
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"entropy": 0.9177370071411133,
|
| 736 |
+
"epoch": 1.5714285714285714,
|
| 737 |
+
"grad_norm": 0.6796875,
|
| 738 |
+
"learning_rate": 2.669481281701739e-05,
|
| 739 |
+
"loss": 0.8557254672050476,
|
| 740 |
+
"mean_token_accuracy": 0.8137136548757553,
|
| 741 |
+
"num_tokens": 1804179.0,
|
| 742 |
+
"step": 69
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"entropy": 0.8276135660707951,
|
| 746 |
+
"epoch": 1.5942857142857143,
|
| 747 |
+
"grad_norm": 0.73828125,
|
| 748 |
+
"learning_rate": 2.420282768545469e-05,
|
| 749 |
+
"loss": 0.6800186634063721,
|
| 750 |
+
"mean_token_accuracy": 0.8346021547913551,
|
| 751 |
+
"num_tokens": 1831612.0,
|
| 752 |
+
"step": 70
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"entropy": 0.9364416636526585,
|
| 756 |
+
"epoch": 1.617142857142857,
|
| 757 |
+
"grad_norm": 0.75390625,
|
| 758 |
+
"learning_rate": 2.181685175319702e-05,
|
| 759 |
+
"loss": 0.7834956049919128,
|
| 760 |
+
"mean_token_accuracy": 0.7997728437185287,
|
| 761 |
+
"num_tokens": 1853595.0,
|
| 762 |
+
"step": 71
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"entropy": 0.8710121884942055,
|
| 766 |
+
"epoch": 1.6400000000000001,
|
| 767 |
+
"grad_norm": 0.765625,
|
| 768 |
+
"learning_rate": 1.9540222023333166e-05,
|
| 769 |
+
"loss": 0.6057661771774292,
|
| 770 |
+
"mean_token_accuracy": 0.8205794207751751,
|
| 771 |
+
"num_tokens": 1882408.0,
|
| 772 |
+
"step": 72
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"entropy": 0.8379442691802979,
|
| 776 |
+
"epoch": 1.6628571428571428,
|
| 777 |
+
"grad_norm": 0.7734375,
|
| 778 |
+
"learning_rate": 1.7376122568400532e-05,
|
| 779 |
+
"loss": 0.7106757760047913,
|
| 780 |
+
"mean_token_accuracy": 0.8180459216237068,
|
| 781 |
+
"num_tokens": 1904944.0,
|
| 782 |
+
"step": 73
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"entropy": 0.8745108097791672,
|
| 786 |
+
"epoch": 1.6857142857142857,
|
| 787 |
+
"grad_norm": 1.2109375,
|
| 788 |
+
"learning_rate": 1.5327580077171587e-05,
|
| 789 |
+
"loss": 0.9369223713874817,
|
| 790 |
+
"mean_token_accuracy": 0.8044208958745003,
|
| 791 |
+
"num_tokens": 1933402.0,
|
| 792 |
+
"step": 74
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"entropy": 0.8298410475254059,
|
| 796 |
+
"epoch": 1.7085714285714286,
|
| 797 |
+
"grad_norm": 0.8828125,
|
| 798 |
+
"learning_rate": 1.339745962155613e-05,
|
| 799 |
+
"loss": 0.6346684694290161,
|
| 800 |
+
"mean_token_accuracy": 0.8335375674068928,
|
| 801 |
+
"num_tokens": 1960093.0,
|
| 802 |
+
"step": 75
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"entropy": 0.8175692446529865,
|
| 806 |
+
"epoch": 1.7314285714285713,
|
| 807 |
+
"grad_norm": 0.91796875,
|
| 808 |
+
"learning_rate": 1.1588460649539035e-05,
|
| 809 |
+
"loss": 0.9119294285774231,
|
| 810 |
+
"mean_token_accuracy": 0.8193050436675549,
|
| 811 |
+
"num_tokens": 1986339.0,
|
| 812 |
+
"step": 76
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"entropy": 0.8975704163312912,
|
| 816 |
+
"epoch": 1.7542857142857144,
|
| 817 |
+
"grad_norm": 0.859375,
|
| 818 |
+
"learning_rate": 9.903113209758096e-06,
|
| 819 |
+
"loss": 0.7237526178359985,
|
| 820 |
+
"mean_token_accuracy": 0.7882252000272274,
|
| 821 |
+
"num_tokens": 2011176.0,
|
| 822 |
+
"step": 77
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"entropy": 0.8363670147955418,
|
| 826 |
+
"epoch": 1.7771428571428571,
|
| 827 |
+
"grad_norm": 0.74609375,
|
| 828 |
+
"learning_rate": 8.343774413002381e-06,
|
| 829 |
+
"loss": 0.6269868016242981,
|
| 830 |
+
"mean_token_accuracy": 0.8103100657463074,
|
| 831 |
+
"num_tokens": 2041289.0,
|
| 832 |
+
"step": 78
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"entropy": 0.8557109721004963,
|
| 836 |
+
"epoch": 1.8,
|
| 837 |
+
"grad_norm": 0.6171875,
|
| 838 |
+
"learning_rate": 6.9126251355795864e-06,
|
| 839 |
+
"loss": 0.5580326914787292,
|
| 840 |
+
"mean_token_accuracy": 0.8539069853723049,
|
| 841 |
+
"num_tokens": 2069418.0,
|
| 842 |
+
"step": 79
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"entropy": 0.790475644171238,
|
| 846 |
+
"epoch": 1.822857142857143,
|
| 847 |
+
"grad_norm": 0.984375,
|
| 848 |
+
"learning_rate": 5.611666969163243e-06,
|
| 849 |
+
"loss": 0.6902757287025452,
|
| 850 |
+
"mean_token_accuracy": 0.8313734903931618,
|
| 851 |
+
"num_tokens": 2101162.0,
|
| 852 |
+
"step": 80
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"entropy": 0.8487820513546467,
|
| 856 |
+
"epoch": 1.8457142857142856,
|
| 857 |
+
"grad_norm": 0.7265625,
|
| 858 |
+
"learning_rate": 4.442719421385922e-06,
|
| 859 |
+
"loss": 0.6248161792755127,
|
| 860 |
+
"mean_token_accuracy": 0.8196811378002167,
|
| 861 |
+
"num_tokens": 2129262.0,
|
| 862 |
+
"step": 81
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"entropy": 0.8948404788970947,
|
| 866 |
+
"epoch": 1.8685714285714285,
|
| 867 |
+
"grad_norm": 0.87890625,
|
| 868 |
+
"learning_rate": 3.40741737109318e-06,
|
| 869 |
+
"loss": 0.8848119378089905,
|
| 870 |
+
"mean_token_accuracy": 0.7821721211075783,
|
| 871 |
+
"num_tokens": 2152597.0,
|
| 872 |
+
"step": 82
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"entropy": 0.9239097908139229,
|
| 876 |
+
"epoch": 1.8914285714285715,
|
| 877 |
+
"grad_norm": 0.7421875,
|
| 878 |
+
"learning_rate": 2.5072087818176382e-06,
|
| 879 |
+
"loss": 0.7065685987472534,
|
| 880 |
+
"mean_token_accuracy": 0.807807132601738,
|
| 881 |
+
"num_tokens": 2177161.0,
|
| 882 |
+
"step": 83
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"entropy": 0.7839450426399708,
|
| 886 |
+
"epoch": 1.9142857142857141,
|
| 887 |
+
"grad_norm": 0.77734375,
|
| 888 |
+
"learning_rate": 1.7433526766711728e-06,
|
| 889 |
+
"loss": 0.7359437942504883,
|
| 890 |
+
"mean_token_accuracy": 0.8403091952204704,
|
| 891 |
+
"num_tokens": 2209858.0,
|
| 892 |
+
"step": 84
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"entropy": 0.843034666031599,
|
| 896 |
+
"epoch": 1.9371428571428573,
|
| 897 |
+
"grad_norm": 0.765625,
|
| 898 |
+
"learning_rate": 1.1169173774871478e-06,
|
| 899 |
+
"loss": 0.7045976519584656,
|
| 900 |
+
"mean_token_accuracy": 0.7968335002660751,
|
| 901 |
+
"num_tokens": 2238762.0,
|
| 902 |
+
"step": 85
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"entropy": 0.8172502107918262,
|
| 906 |
+
"epoch": 1.96,
|
| 907 |
+
"grad_norm": 0.73046875,
|
| 908 |
+
"learning_rate": 6.287790106757396e-07,
|
| 909 |
+
"loss": 0.7153176665306091,
|
| 910 |
+
"mean_token_accuracy": 0.8268864750862122,
|
| 911 |
+
"num_tokens": 2266000.0,
|
| 912 |
+
"step": 86
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"entropy": 0.9393381252884865,
|
| 916 |
+
"epoch": 1.9828571428571429,
|
| 917 |
+
"grad_norm": 0.8046875,
|
| 918 |
+
"learning_rate": 2.7962028188198706e-07,
|
| 919 |
+
"loss": 0.7222880721092224,
|
| 920 |
+
"mean_token_accuracy": 0.8043639399111271,
|
| 921 |
+
"num_tokens": 2291896.0,
|
| 922 |
+
"step": 87
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"entropy": 0.9007064700126648,
|
| 926 |
+
"epoch": 2.0,
|
| 927 |
+
"grad_norm": 0.9453125,
|
| 928 |
+
"learning_rate": 6.992952116013918e-08,
|
| 929 |
+
"loss": 0.9412174224853516,
|
| 930 |
+
"mean_token_accuracy": 0.7673612783352534,
|
| 931 |
+
"num_tokens": 2307166.0,
|
| 932 |
+
"step": 88
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.0,
|
| 936 |
+
"eval_entropy": 0.8574838742613793,
|
| 937 |
+
"eval_loss": 0.9815174341201782,
|
| 938 |
+
"eval_mean_token_accuracy": 0.7545835164189338,
|
| 939 |
+
"eval_num_tokens": 2307166.0,
|
| 940 |
+
"eval_runtime": 77.2359,
|
| 941 |
+
"eval_samples_per_second": 1.295,
|
| 942 |
+
"eval_steps_per_second": 1.295,
|
| 943 |
+
"step": 88
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 2.0,
|
| 947 |
+
"step": 88,
|
| 948 |
+
"total_flos": 1.0597781694661018e+17,
|
| 949 |
+
"train_loss": 0.9580497294664383,
|
| 950 |
+
"train_runtime": 4168.6102,
|
| 951 |
+
"train_samples_per_second": 0.336,
|
| 952 |
+
"train_steps_per_second": 0.021
|
| 953 |
+
}
|
| 954 |
+
],
|
| 955 |
+
"trial_params": null,
|
| 956 |
+
"global_step": 88,
|
| 957 |
+
"is_local_process_zero": true,
|
| 958 |
+
"total_flos": 1.0597781694661018e+17,
|
| 959 |
+
"num_train_epochs": 2,
|
| 960 |
+
"epoch": 2.0,
|
| 961 |
+
"best_metric": null,
|
| 962 |
+
"trial_name": null
|
| 963 |
+
}
|