Text Generation
Transformers
Safetensors
English
testgeniy
causal-lm
reasoning
mathematics
logic
long-context
4k-context
small-language-model
Instructions to use Asilarkness/testgeniy with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Asilarkness/testgeniy with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Asilarkness/testgeniy")# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("Asilarkness/testgeniy", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Asilarkness/testgeniy with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Asilarkness/testgeniy" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Asilarkness/testgeniy", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/Asilarkness/testgeniy
- SGLang
How to use Asilarkness/testgeniy with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Asilarkness/testgeniy" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Asilarkness/testgeniy", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Asilarkness/testgeniy" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Asilarkness/testgeniy", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use Asilarkness/testgeniy with Docker Model Runner:
docker model run hf.co/Asilarkness/testgeniy
Add reasoning SFT v3 training report
Browse files
candidates/budgie-alignment-v2/reasoning-sft-then-rl-v3/recovery/v3_sft_training_report.json
ADDED
|
@@ -0,0 +1,1833 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"method": "one-epoch long verified reasoning SFT with sampled-action leader KL",
|
| 3 |
+
"base": "verified-math-a025",
|
| 4 |
+
"encoded_rows": 908,
|
| 5 |
+
"math_rows": 800,
|
| 6 |
+
"info_rows": 108,
|
| 7 |
+
"updates": 227,
|
| 8 |
+
"grad_accum": 4,
|
| 9 |
+
"peak_lr": 3e-07,
|
| 10 |
+
"kl_beta": 0.12,
|
| 11 |
+
"max_length": 1536,
|
| 12 |
+
"history": [
|
| 13 |
+
{
|
| 14 |
+
"update": 1,
|
| 15 |
+
"ce": 3.538551688194275,
|
| 16 |
+
"kl": 0.001223451952682808,
|
| 17 |
+
"drift_frac": 0.0842948742210865,
|
| 18 |
+
"grad_norm": 97.95021057128906,
|
| 19 |
+
"lr": 2.5e-08
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"update": 2,
|
| 23 |
+
"ce": 3.7247090339660645,
|
| 24 |
+
"kl": 0.0011211548408027738,
|
| 25 |
+
"drift_frac": 0.05625000037252903,
|
| 26 |
+
"grad_norm": 132.21041870117188,
|
| 27 |
+
"lr": 5e-08
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"update": 3,
|
| 31 |
+
"ce": 2.2506245374679565,
|
| 32 |
+
"kl": 0.0021406779997050762,
|
| 33 |
+
"drift_frac": 0.14215686544775963,
|
| 34 |
+
"grad_norm": 198.66062927246094,
|
| 35 |
+
"lr": 7.5e-08
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"update": 4,
|
| 39 |
+
"ce": 2.4903512001037598,
|
| 40 |
+
"kl": 0.0013115592009853572,
|
| 41 |
+
"drift_frac": 0.07763157971203327,
|
| 42 |
+
"grad_norm": 134.6076202392578,
|
| 43 |
+
"lr": 1e-07
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"update": 5,
|
| 47 |
+
"ce": 3.9587668776512146,
|
| 48 |
+
"kl": 0.0013989002327434719,
|
| 49 |
+
"drift_frac": 0.06827731523662806,
|
| 50 |
+
"grad_norm": 104.69552612304688,
|
| 51 |
+
"lr": 1.25e-07
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"update": 6,
|
| 55 |
+
"ce": 4.269447326660156,
|
| 56 |
+
"kl": 0.0015167651654337533,
|
| 57 |
+
"drift_frac": 0.08613445702940226,
|
| 58 |
+
"grad_norm": 116.19329071044922,
|
| 59 |
+
"lr": 1.5e-07
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"update": 7,
|
| 63 |
+
"ce": 2.496525228023529,
|
| 64 |
+
"kl": 0.0008211426757043228,
|
| 65 |
+
"drift_frac": 0.029411764815449715,
|
| 66 |
+
"grad_norm": 105.23081970214844,
|
| 67 |
+
"lr": 1.75e-07
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"update": 8,
|
| 71 |
+
"ce": 2.5901973843574524,
|
| 72 |
+
"kl": 0.0018725720728980377,
|
| 73 |
+
"drift_frac": 0.07638888899236917,
|
| 74 |
+
"grad_norm": 97.65045928955078,
|
| 75 |
+
"lr": 2e-07
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"update": 9,
|
| 79 |
+
"ce": 3.5735539197921753,
|
| 80 |
+
"kl": 0.0017626602930249646,
|
| 81 |
+
"drift_frac": 0.0625,
|
| 82 |
+
"grad_norm": 359.466064453125,
|
| 83 |
+
"lr": 2.25e-07
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"update": 10,
|
| 87 |
+
"ce": 4.108733654022217,
|
| 88 |
+
"kl": 0.005274100694805384,
|
| 89 |
+
"drift_frac": 0.12916667014360428,
|
| 90 |
+
"grad_norm": 251.59793090820312,
|
| 91 |
+
"lr": 2.5e-07
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"update": 11,
|
| 95 |
+
"ce": 2.626381754875183,
|
| 96 |
+
"kl": 0.0018866165482904762,
|
| 97 |
+
"drift_frac": 0.10835913009941578,
|
| 98 |
+
"grad_norm": 208.23765563964844,
|
| 99 |
+
"lr": 2.7499999999999996e-07
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"update": 12,
|
| 103 |
+
"ce": 2.2875999808311462,
|
| 104 |
+
"kl": 0.001339451817329973,
|
| 105 |
+
"drift_frac": 0.06227106414735317,
|
| 106 |
+
"grad_norm": 69.71000671386719,
|
| 107 |
+
"lr": 3e-07
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"update": 13,
|
| 111 |
+
"ce": 5.156558513641357,
|
| 112 |
+
"kl": 0.0030864934669807553,
|
| 113 |
+
"drift_frac": 0.09307359531521797,
|
| 114 |
+
"grad_norm": 240.46498107910156,
|
| 115 |
+
"lr": 3e-07
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"update": 14,
|
| 119 |
+
"ce": 3.425640344619751,
|
| 120 |
+
"kl": 0.000992244080407545,
|
| 121 |
+
"drift_frac": 0.0714285746216774,
|
| 122 |
+
"grad_norm": 92.13054656982422,
|
| 123 |
+
"lr": 2.999854531764202e-07
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"update": 15,
|
| 127 |
+
"ce": 3.8330267667770386,
|
| 128 |
+
"kl": 0.001204864980536513,
|
| 129 |
+
"drift_frac": 0.09375,
|
| 130 |
+
"grad_norm": 100.53870391845703,
|
| 131 |
+
"lr": 2.999418158406448e-07
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"update": 16,
|
| 135 |
+
"ce": 3.0005993843078613,
|
| 136 |
+
"kl": 0.0012357652885839343,
|
| 137 |
+
"drift_frac": 0.09699755255132914,
|
| 138 |
+
"grad_norm": 44.343238830566406,
|
| 139 |
+
"lr": 2.998690973968905e-07
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"update": 17,
|
| 143 |
+
"ce": 2.8656360507011414,
|
| 144 |
+
"kl": 0.0012986478905077092,
|
| 145 |
+
"drift_frac": 0.05488445423543453,
|
| 146 |
+
"grad_norm": 66.21858978271484,
|
| 147 |
+
"lr": 2.9976731351659987e-07
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"update": 18,
|
| 151 |
+
"ce": 3.1336140036582947,
|
| 152 |
+
"kl": 0.0010291965445503592,
|
| 153 |
+
"drift_frac": 0.040178571827709675,
|
| 154 |
+
"grad_norm": 76.4161376953125,
|
| 155 |
+
"lr": 2.99636486135064e-07
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"update": 19,
|
| 159 |
+
"ce": 3.640356719493866,
|
| 160 |
+
"kl": 0.0007088115089572966,
|
| 161 |
+
"drift_frac": 0.02380952425301075,
|
| 162 |
+
"grad_norm": 100.51805114746094,
|
| 163 |
+
"lr": 2.994766434466953e-07
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"update": 20,
|
| 167 |
+
"ce": 3.6283442080020905,
|
| 168 |
+
"kl": 0.0024094453547149897,
|
| 169 |
+
"drift_frac": 0.16437729634344578,
|
| 170 |
+
"grad_norm": 80.80867767333984,
|
| 171 |
+
"lr": 2.9928781989895147e-07
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"update": 21,
|
| 175 |
+
"ce": 3.5531338453292847,
|
| 176 |
+
"kl": 0.0006767813538317569,
|
| 177 |
+
"drift_frac": 0.02884615585207939,
|
| 178 |
+
"grad_norm": 56.63362121582031,
|
| 179 |
+
"lr": 2.9907005618491146e-07
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"update": 22,
|
| 183 |
+
"ce": 2.6485652327537537,
|
| 184 |
+
"kl": 0.0006122057457105257,
|
| 185 |
+
"drift_frac": 0.03203253913670778,
|
| 186 |
+
"grad_norm": 59.86662292480469,
|
| 187 |
+
"lr": 2.9882339923450607e-07
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"update": 23,
|
| 191 |
+
"ce": 3.031906545162201,
|
| 192 |
+
"kl": 0.00075152212230023,
|
| 193 |
+
"drift_frac": 0.029411764815449715,
|
| 194 |
+
"grad_norm": 68.89799499511719,
|
| 195 |
+
"lr": 2.985479022044042e-07
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"update": 24,
|
| 199 |
+
"ce": 3.317367911338806,
|
| 200 |
+
"kl": 0.0006392341165337712,
|
| 201 |
+
"drift_frac": 0.02380952425301075,
|
| 202 |
+
"grad_norm": 72.74163818359375,
|
| 203 |
+
"lr": 2.982436244665568e-07
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"update": 25,
|
| 207 |
+
"ce": 3.539289712905884,
|
| 208 |
+
"kl": 0.001575481059262529,
|
| 209 |
+
"drift_frac": 0.0714285746216774,
|
| 210 |
+
"grad_norm": 54.52531433105469,
|
| 211 |
+
"lr": 2.979106315954019e-07
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"update": 26,
|
| 215 |
+
"ce": 3.5921270847320557,
|
| 216 |
+
"kl": 0.0008827472047414631,
|
| 217 |
+
"drift_frac": 0.03333333507180214,
|
| 218 |
+
"grad_norm": 88.11039733886719,
|
| 219 |
+
"lr": 2.9754899535373294e-07
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"update": 27,
|
| 223 |
+
"ce": 4.183539390563965,
|
| 224 |
+
"kl": 0.002558601088821888,
|
| 225 |
+
"drift_frac": 0.12857143208384514,
|
| 226 |
+
"grad_norm": 146.65953063964844,
|
| 227 |
+
"lr": 2.97158793677233e-07
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"update": 28,
|
| 231 |
+
"ce": 1.325046181678772,
|
| 232 |
+
"kl": 0.0010006816883105785,
|
| 233 |
+
"drift_frac": 0.06299429945647717,
|
| 234 |
+
"grad_norm": 11.920963287353516,
|
| 235 |
+
"lr": 2.9674011065767897e-07
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"update": 29,
|
| 239 |
+
"ce": 1.403570979833603,
|
| 240 |
+
"kl": 0.0009888321219477803,
|
| 241 |
+
"drift_frac": 0.052271944005042315,
|
| 242 |
+
"grad_norm": 9.721692085266113,
|
| 243 |
+
"lr": 2.962930365248194e-07
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"update": 30,
|
| 247 |
+
"ce": 1.1523630321025848,
|
| 248 |
+
"kl": 0.0009369849649374373,
|
| 249 |
+
"drift_frac": 0.04682533466257155,
|
| 250 |
+
"grad_norm": 7.417837619781494,
|
| 251 |
+
"lr": 2.9581766762692897e-07
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"update": 31,
|
| 255 |
+
"ce": 1.4184409081935883,
|
| 256 |
+
"kl": 0.0007759521249681711,
|
| 257 |
+
"drift_frac": 0.04882436478510499,
|
| 258 |
+
"grad_norm": 9.458812713623047,
|
| 259 |
+
"lr": 2.9531410641004455e-07
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"update": 32,
|
| 263 |
+
"ce": 1.1053877770900726,
|
| 264 |
+
"kl": 0.0007806056382833049,
|
| 265 |
+
"drift_frac": 0.04721762053668499,
|
| 266 |
+
"grad_norm": 8.717150688171387,
|
| 267 |
+
"lr": 2.947824613958874e-07
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"update": 33,
|
| 271 |
+
"ce": 1.2606067508459091,
|
| 272 |
+
"kl": 0.00072369264671579,
|
| 273 |
+
"drift_frac": 0.051310960203409195,
|
| 274 |
+
"grad_norm": 8.279505729675293,
|
| 275 |
+
"lr": 2.942228471584757e-07
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"update": 34,
|
| 279 |
+
"ce": 1.6003108024597168,
|
| 280 |
+
"kl": 0.0008679848033352755,
|
| 281 |
+
"drift_frac": 0.04033056506887078,
|
| 282 |
+
"grad_norm": 10.196907043457031,
|
| 283 |
+
"lr": 2.9363538429943293e-07
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"update": 35,
|
| 287 |
+
"ce": 1.593827337026596,
|
| 288 |
+
"kl": 0.000813726699561812,
|
| 289 |
+
"drift_frac": 0.03967779455706477,
|
| 290 |
+
"grad_norm": 14.789701461791992,
|
| 291 |
+
"lr": 2.9302019942199715e-07
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"update": 36,
|
| 295 |
+
"ce": 1.2804492115974426,
|
| 296 |
+
"kl": 0.0007756019767839462,
|
| 297 |
+
"drift_frac": 0.04920882638543844,
|
| 298 |
+
"grad_norm": 7.81544303894043,
|
| 299 |
+
"lr": 2.9237742510373703e-07
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"update": 37,
|
| 303 |
+
"ce": 1.2322860807180405,
|
| 304 |
+
"kl": 0.000988430532743223,
|
| 305 |
+
"drift_frac": 0.06842744257301092,
|
| 306 |
+
"grad_norm": 7.467689037322998,
|
| 307 |
+
"lr": 2.9170719986798003e-07
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"update": 38,
|
| 311 |
+
"ce": 1.229136049747467,
|
| 312 |
+
"kl": 0.0008774271555012092,
|
| 313 |
+
"drift_frac": 0.05900773126631975,
|
| 314 |
+
"grad_norm": 8.759092330932617,
|
| 315 |
+
"lr": 2.9100966815395964e-07
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"update": 39,
|
| 319 |
+
"ce": 1.300631731748581,
|
| 320 |
+
"kl": 0.0006411699505406432,
|
| 321 |
+
"drift_frac": 0.034251566510647535,
|
| 322 |
+
"grad_norm": 8.404024124145508,
|
| 323 |
+
"lr": 2.902849802856877e-07
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"update": 40,
|
| 327 |
+
"ce": 1.132405772805214,
|
| 328 |
+
"kl": 0.0008723139180801809,
|
| 329 |
+
"drift_frac": 0.04914182750508189,
|
| 330 |
+
"grad_norm": 6.600926876068115,
|
| 331 |
+
"lr": 2.8953329243955764e-07
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"update": 41,
|
| 335 |
+
"ce": 1.4244795143604279,
|
| 336 |
+
"kl": 0.0008519404073012993,
|
| 337 |
+
"drift_frac": 0.058619968593120575,
|
| 338 |
+
"grad_norm": 7.896564960479736,
|
| 339 |
+
"lr": 2.887547666106879e-07
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"update": 42,
|
| 343 |
+
"ce": 1.2710935026407242,
|
| 344 |
+
"kl": 0.0009223220404237509,
|
| 345 |
+
"drift_frac": 0.05881678685545921,
|
| 346 |
+
"grad_norm": 10.9074068069458,
|
| 347 |
+
"lr": 2.8794957057801e-07
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"update": 43,
|
| 351 |
+
"ce": 1.3473848700523376,
|
| 352 |
+
"kl": 0.0007307154446607456,
|
| 353 |
+
"drift_frac": 0.04500186396762729,
|
| 354 |
+
"grad_norm": 7.803529739379883,
|
| 355 |
+
"lr": 2.871178778681111e-07
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"update": 44,
|
| 359 |
+
"ce": 1.2855714708566666,
|
| 360 |
+
"kl": 0.0005734634032705799,
|
| 361 |
+
"drift_frac": 0.03194753034040332,
|
| 362 |
+
"grad_norm": 7.435911178588867,
|
| 363 |
+
"lr": 2.862598677178376e-07
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"update": 45,
|
| 367 |
+
"ce": 1.745590716600418,
|
| 368 |
+
"kl": 0.00089225810370408,
|
| 369 |
+
"drift_frac": 0.0430073207244277,
|
| 370 |
+
"grad_norm": 13.365913391113281,
|
| 371 |
+
"lr": 2.853757250356677e-07
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"update": 46,
|
| 375 |
+
"ce": 1.2935452312231064,
|
| 376 |
+
"kl": 0.0007396720684482716,
|
| 377 |
+
"drift_frac": 0.04109058855101466,
|
| 378 |
+
"grad_norm": 7.639344215393066,
|
| 379 |
+
"lr": 2.8446564036186245e-07
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"update": 47,
|
| 383 |
+
"ce": 1.403597429394722,
|
| 384 |
+
"kl": 0.000791062688222155,
|
| 385 |
+
"drift_frac": 0.04710930911824107,
|
| 386 |
+
"grad_norm": 8.35036849975586,
|
| 387 |
+
"lr": 2.835298098274025e-07
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"update": 48,
|
| 391 |
+
"ce": 1.2043190896511078,
|
| 392 |
+
"kl": 0.0006641525396844372,
|
| 393 |
+
"drift_frac": 0.041309695690870285,
|
| 394 |
+
"grad_norm": 7.528050422668457,
|
| 395 |
+
"lr": 2.825684351117201e-07
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"update": 49,
|
| 399 |
+
"ce": 1.1160235852003098,
|
| 400 |
+
"kl": 0.0006896877384861,
|
| 401 |
+
"drift_frac": 0.03284739563241601,
|
| 402 |
+
"grad_norm": 7.1364874839782715,
|
| 403 |
+
"lr": 2.815817233992358e-07
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"update": 50,
|
| 407 |
+
"ce": 1.5810885727405548,
|
| 408 |
+
"kl": 0.0008001479291124269,
|
| 409 |
+
"drift_frac": 0.04724500887095928,
|
| 410 |
+
"grad_norm": 7.52408504486084,
|
| 411 |
+
"lr": 2.8056988733470834e-07
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"update": 51,
|
| 415 |
+
"ce": 1.2303399443626404,
|
| 416 |
+
"kl": 0.0007196556689450517,
|
| 417 |
+
"drift_frac": 0.03665072191506624,
|
| 418 |
+
"grad_norm": 6.465488910675049,
|
| 419 |
+
"lr": 2.795331449774075e-07
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"update": 52,
|
| 423 |
+
"ce": 1.3189209997653961,
|
| 424 |
+
"kl": 0.0007431682606693357,
|
| 425 |
+
"drift_frac": 0.04172079684212804,
|
| 426 |
+
"grad_norm": 7.317380428314209,
|
| 427 |
+
"lr": 2.784717197541211e-07
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"update": 53,
|
| 431 |
+
"ce": 1.3691906332969666,
|
| 432 |
+
"kl": 0.0008672039402881637,
|
| 433 |
+
"drift_frac": 0.05027059093117714,
|
| 434 |
+
"grad_norm": 6.821427345275879,
|
| 435 |
+
"lr": 2.77385840411004e-07
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"update": 54,
|
| 439 |
+
"ce": 1.1744352728128433,
|
| 440 |
+
"kl": 0.0005493841963470913,
|
| 441 |
+
"drift_frac": 0.023992398288100958,
|
| 442 |
+
"grad_norm": 11.350872039794922,
|
| 443 |
+
"lr": 2.762757409642822e-07
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"update": 55,
|
| 447 |
+
"ce": 0.9146636575460434,
|
| 448 |
+
"kl": 0.0007157866930356249,
|
| 449 |
+
"drift_frac": 0.04177132062613964,
|
| 450 |
+
"grad_norm": 5.519471645355225,
|
| 451 |
+
"lr": 2.7514166064981917e-07
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"update": 56,
|
| 455 |
+
"ce": 1.2082852125167847,
|
| 456 |
+
"kl": 0.0008967171743279323,
|
| 457 |
+
"drift_frac": 0.057539583183825016,
|
| 458 |
+
"grad_norm": 6.895761489868164,
|
| 459 |
+
"lr": 2.7398384387155983e-07
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"update": 57,
|
| 463 |
+
"ce": 1.4530055224895477,
|
| 464 |
+
"kl": 0.0009128262900048867,
|
| 465 |
+
"drift_frac": 0.051596757024526596,
|
| 466 |
+
"grad_norm": 7.215538501739502,
|
| 467 |
+
"lr": 2.728025401488581e-07
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"update": 58,
|
| 471 |
+
"ce": 0.8601672947406769,
|
| 472 |
+
"kl": 0.0006162677746033296,
|
| 473 |
+
"drift_frac": 0.038331746589392424,
|
| 474 |
+
"grad_norm": 5.206908702850342,
|
| 475 |
+
"lr": 2.7159800406270456e-07
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"update": 59,
|
| 479 |
+
"ce": 1.4296726733446121,
|
| 480 |
+
"kl": 0.0008538641413906589,
|
| 481 |
+
"drift_frac": 0.05090134497731924,
|
| 482 |
+
"grad_norm": 6.5874223709106445,
|
| 483 |
+
"lr": 2.7037049520086116e-07
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"update": 60,
|
| 487 |
+
"ce": 1.069526955485344,
|
| 488 |
+
"kl": 0.0006327034498099238,
|
| 489 |
+
"drift_frac": 0.03250848315656185,
|
| 490 |
+
"grad_norm": 6.299490928649902,
|
| 491 |
+
"lr": 2.6912027810191846e-07
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"update": 61,
|
| 495 |
+
"ce": 1.2052197605371475,
|
| 496 |
+
"kl": 0.000798501649114769,
|
| 497 |
+
"drift_frac": 0.04450059495866299,
|
| 498 |
+
"grad_norm": 6.234531879425049,
|
| 499 |
+
"lr": 2.678476221982849e-07
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"update": 62,
|
| 503 |
+
"ce": 1.090393990278244,
|
| 504 |
+
"kl": 0.0008333805744769052,
|
| 505 |
+
"drift_frac": 0.046888030134141445,
|
| 506 |
+
"grad_norm": 6.188854217529297,
|
| 507 |
+
"lr": 2.665528017581222e-07
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"update": 63,
|
| 511 |
+
"ce": 1.1771894842386246,
|
| 512 |
+
"kl": 0.000885220113559626,
|
| 513 |
+
"drift_frac": 0.04329330660402775,
|
| 514 |
+
"grad_norm": 5.898693084716797,
|
| 515 |
+
"lr": 2.6523609582623774e-07
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"update": 64,
|
| 519 |
+
"ce": 0.9992333799600601,
|
| 520 |
+
"kl": 0.0008020212408155203,
|
| 521 |
+
"drift_frac": 0.05296085122972727,
|
| 522 |
+
"grad_norm": 5.337985992431641,
|
| 523 |
+
"lr": 2.638977881639486e-07
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"update": 65,
|
| 527 |
+
"ce": 1.1238050609827042,
|
| 528 |
+
"kl": 0.0006481189775513485,
|
| 529 |
+
"drift_frac": 0.026658230926841497,
|
| 530 |
+
"grad_norm": 6.046208381652832,
|
| 531 |
+
"lr": 2.6253816718792833e-07
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"update": 66,
|
| 535 |
+
"ce": 1.2460280060768127,
|
| 536 |
+
"kl": 0.000899813836440444,
|
| 537 |
+
"drift_frac": 0.04985949443653226,
|
| 538 |
+
"grad_norm": 5.792781829833984,
|
| 539 |
+
"lr": 2.6115752590805046e-07
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"update": 67,
|
| 543 |
+
"ce": 1.4626020193099976,
|
| 544 |
+
"kl": 0.000941945516387932,
|
| 545 |
+
"drift_frac": 0.04764819983392954,
|
| 546 |
+
"grad_norm": 8.468130111694336,
|
| 547 |
+
"lr": 2.597561618642429e-07
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"update": 68,
|
| 551 |
+
"ce": 1.4544167220592499,
|
| 552 |
+
"kl": 0.000894820477697067,
|
| 553 |
+
"drift_frac": 0.04709818586707115,
|
| 554 |
+
"grad_norm": 8.969865798950195,
|
| 555 |
+
"lr": 2.583343770623651e-07
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"update": 69,
|
| 559 |
+
"ce": 1.3955807089805603,
|
| 560 |
+
"kl": 0.000805779782240279,
|
| 561 |
+
"drift_frac": 0.04070504195988178,
|
| 562 |
+
"grad_norm": 6.130509376525879,
|
| 563 |
+
"lr": 2.568924779091235e-07
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"update": 70,
|
| 567 |
+
"ce": 1.025430127978325,
|
| 568 |
+
"kl": 0.0008046418661251664,
|
| 569 |
+
"drift_frac": 0.045382329262793064,
|
| 570 |
+
"grad_norm": 5.295656681060791,
|
| 571 |
+
"lr": 2.55430775146038e-07
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"update": 71,
|
| 575 |
+
"ce": 0.9171223789453506,
|
| 576 |
+
"kl": 0.0008278430032078177,
|
| 577 |
+
"drift_frac": 0.04534016828984022,
|
| 578 |
+
"grad_norm": 4.571761608123779,
|
| 579 |
+
"lr": 2.5394958378247495e-07
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"update": 72,
|
| 583 |
+
"ce": 1.2810973674058914,
|
| 584 |
+
"kl": 0.0007998075452633202,
|
| 585 |
+
"drift_frac": 0.048564073629677296,
|
| 586 |
+
"grad_norm": 5.635980129241943,
|
| 587 |
+
"lr": 2.524492230277596e-07
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"update": 73,
|
| 591 |
+
"ce": 1.2809077203273773,
|
| 592 |
+
"kl": 0.0007367939833784476,
|
| 593 |
+
"drift_frac": 0.04585993941873312,
|
| 594 |
+
"grad_norm": 6.78187370300293,
|
| 595 |
+
"lr": 2.5093001622238383e-07
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"update": 74,
|
| 599 |
+
"ce": 1.3074978590011597,
|
| 600 |
+
"kl": 0.0010324735048925504,
|
| 601 |
+
"drift_frac": 0.056532954797148705,
|
| 602 |
+
"grad_norm": 5.791697025299072,
|
| 603 |
+
"lr": 2.493922907683236e-07
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"update": 75,
|
| 607 |
+
"ce": 1.514907419681549,
|
| 608 |
+
"kl": 0.0008105379238259047,
|
| 609 |
+
"drift_frac": 0.0467648645862937,
|
| 610 |
+
"grad_norm": 5.778327465057373,
|
| 611 |
+
"lr": 2.478363780584812e-07
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"update": 76,
|
| 615 |
+
"ce": 1.098328024148941,
|
| 616 |
+
"kl": 0.0008163142629200593,
|
| 617 |
+
"drift_frac": 0.05169339291751385,
|
| 618 |
+
"grad_norm": 5.003957748413086,
|
| 619 |
+
"lr": 2.4626261340526666e-07
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"update": 77,
|
| 623 |
+
"ce": 1.3024695217609406,
|
| 624 |
+
"kl": 0.0009096241847146302,
|
| 625 |
+
"drift_frac": 0.051626091822981834,
|
| 626 |
+
"grad_norm": 5.3108978271484375,
|
| 627 |
+
"lr": 2.4467133596833606e-07
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"update": 78,
|
| 631 |
+
"ce": 1.0044908598065376,
|
| 632 |
+
"kl": 0.0006562793714692816,
|
| 633 |
+
"drift_frac": 0.03572519589215517,
|
| 634 |
+
"grad_norm": 4.71830415725708,
|
| 635 |
+
"lr": 2.4306288868149873e-07
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"update": 79,
|
| 639 |
+
"ce": 1.301384523510933,
|
| 640 |
+
"kl": 0.0008808651327854022,
|
| 641 |
+
"drift_frac": 0.05123406741768122,
|
| 642 |
+
"grad_norm": 7.11338996887207,
|
| 643 |
+
"lr": 2.414376181788128e-07
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"update": 80,
|
| 647 |
+
"ce": 1.1953087747097015,
|
| 648 |
+
"kl": 0.0008735001902095973,
|
| 649 |
+
"drift_frac": 0.0544233163818717,
|
| 650 |
+
"grad_norm": 4.67734432220459,
|
| 651 |
+
"lr": 2.3979587471988227e-07
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"update": 81,
|
| 655 |
+
"ce": 1.0101770386099815,
|
| 656 |
+
"kl": 0.0009059560688911006,
|
| 657 |
+
"drift_frac": 0.05304779810830951,
|
| 658 |
+
"grad_norm": 4.568741798400879,
|
| 659 |
+
"lr": 2.3813801211437322e-07
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"update": 82,
|
| 663 |
+
"ce": 1.0523600578308105,
|
| 664 |
+
"kl": 0.0008110712369671091,
|
| 665 |
+
"drift_frac": 0.05493324436247349,
|
| 666 |
+
"grad_norm": 4.91206169128418,
|
| 667 |
+
"lr": 2.3646438764576486e-07
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"update": 83,
|
| 671 |
+
"ce": 1.4748104512691498,
|
| 672 |
+
"kl": 0.0010775311529869214,
|
| 673 |
+
"drift_frac": 0.05605477746576071,
|
| 674 |
+
"grad_norm": 6.065121173858643,
|
| 675 |
+
"lr": 2.3477536199435167e-07
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"update": 84,
|
| 679 |
+
"ce": 1.173541247844696,
|
| 680 |
+
"kl": 0.0007839623285690323,
|
| 681 |
+
"drift_frac": 0.040957994759082794,
|
| 682 |
+
"grad_norm": 4.7310051918029785,
|
| 683 |
+
"lr": 2.3307129915951407e-07
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"update": 85,
|
| 687 |
+
"ce": 1.2565110176801682,
|
| 688 |
+
"kl": 0.0009195466118399054,
|
| 689 |
+
"drift_frac": 0.052667141892015934,
|
| 690 |
+
"grad_norm": 4.819740295410156,
|
| 691 |
+
"lr": 2.3135256638127315e-07
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"update": 86,
|
| 695 |
+
"ce": 0.9261077120900154,
|
| 696 |
+
"kl": 0.0006547942248289473,
|
| 697 |
+
"drift_frac": 0.028973964042961597,
|
| 698 |
+
"grad_norm": 4.636392593383789,
|
| 699 |
+
"lr": 2.296195340611475e-07
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"update": 87,
|
| 703 |
+
"ce": 1.165843904018402,
|
| 704 |
+
"kl": 0.0009094122651731595,
|
| 705 |
+
"drift_frac": 0.054654247127473354,
|
| 706 |
+
"grad_norm": 7.7583394050598145,
|
| 707 |
+
"lr": 2.2787257568232857e-07
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"update": 88,
|
| 711 |
+
"ce": 0.6246384084224701,
|
| 712 |
+
"kl": 0.0005384607429732569,
|
| 713 |
+
"drift_frac": 0.031362848123535514,
|
| 714 |
+
"grad_norm": 3.827564001083374,
|
| 715 |
+
"lr": 2.2611206772919172e-07
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"update": 89,
|
| 719 |
+
"ce": 0.8072536587715149,
|
| 720 |
+
"kl": 0.0008932169293984771,
|
| 721 |
+
"drift_frac": 0.053428469225764275,
|
| 722 |
+
"grad_norm": 3.9820549488067627,
|
| 723 |
+
"lr": 2.2433838960616072e-07
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"update": 90,
|
| 727 |
+
"ce": 0.9621523097157478,
|
| 728 |
+
"kl": 0.0007514993958466221,
|
| 729 |
+
"drift_frac": 0.041160945780575275,
|
| 730 |
+
"grad_norm": 3.9328701496124268,
|
| 731 |
+
"lr": 2.2255192355594285e-07
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"update": 91,
|
| 735 |
+
"ce": 1.0217181146144867,
|
| 736 |
+
"kl": 0.0008546207973267883,
|
| 737 |
+
"drift_frac": 0.04648606665432453,
|
| 738 |
+
"grad_norm": 4.333801746368408,
|
| 739 |
+
"lr": 2.2075305457715245e-07
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"update": 92,
|
| 743 |
+
"ce": 0.982524711638689,
|
| 744 |
+
"kl": 0.0006494432964245789,
|
| 745 |
+
"drift_frac": 0.03713762294501066,
|
| 746 |
+
"grad_norm": 5.01400089263916,
|
| 747 |
+
"lr": 2.189421703413402e-07
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"update": 93,
|
| 751 |
+
"ce": 0.823276475071907,
|
| 752 |
+
"kl": 0.0008681819381308742,
|
| 753 |
+
"drift_frac": 0.048628955613821745,
|
| 754 |
+
"grad_norm": 3.8080289363861084,
|
| 755 |
+
"lr": 2.17119661109447e-07
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"update": 94,
|
| 759 |
+
"ce": 1.1629377007484436,
|
| 760 |
+
"kl": 0.0008835747721605003,
|
| 761 |
+
"drift_frac": 0.04425085335969925,
|
| 762 |
+
"grad_norm": 4.81643533706665,
|
| 763 |
+
"lr": 2.15285919647699e-07
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"update": 95,
|
| 767 |
+
"ce": 0.9892039895057678,
|
| 768 |
+
"kl": 0.0009211189826601185,
|
| 769 |
+
"drift_frac": 0.03626193443778902,
|
| 770 |
+
"grad_norm": 5.977481365203857,
|
| 771 |
+
"lr": 2.134413411429634e-07
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"update": 96,
|
| 775 |
+
"ce": 1.3896388113498688,
|
| 776 |
+
"kl": 0.0010265405289828777,
|
| 777 |
+
"drift_frac": 0.05348957143723965,
|
| 778 |
+
"grad_norm": 5.797865867614746,
|
| 779 |
+
"lr": 2.115863231175821e-07
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"update": 97,
|
| 783 |
+
"ce": 0.8354207128286362,
|
| 784 |
+
"kl": 0.0007118851353880018,
|
| 785 |
+
"drift_frac": 0.041735677514225245,
|
| 786 |
+
"grad_norm": 3.7527008056640625,
|
| 787 |
+
"lr": 2.0972126534370277e-07
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"update": 98,
|
| 791 |
+
"ce": 0.9529134035110474,
|
| 792 |
+
"kl": 0.0011562850413611159,
|
| 793 |
+
"drift_frac": 0.05421243514865637,
|
| 794 |
+
"grad_norm": 3.6398327350616455,
|
| 795 |
+
"lr": 2.078465697571238e-07
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"update": 99,
|
| 799 |
+
"ce": 1.1178778260946274,
|
| 800 |
+
"kl": 0.0007249849440995604,
|
| 801 |
+
"drift_frac": 0.03873869776725769,
|
| 802 |
+
"grad_norm": 5.02525520324707,
|
| 803 |
+
"lr": 2.059626403706744e-07
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"update": 100,
|
| 807 |
+
"ce": 0.8577919974923134,
|
| 808 |
+
"kl": 0.0006445020189858042,
|
| 809 |
+
"drift_frac": 0.02807536325417459,
|
| 810 |
+
"grad_norm": 4.021769046783447,
|
| 811 |
+
"lr": 2.0406988318714592e-07
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"update": 101,
|
| 815 |
+
"ce": 0.9204409122467041,
|
| 816 |
+
"kl": 0.0009804619912756607,
|
| 817 |
+
"drift_frac": 0.046440054662525654,
|
| 818 |
+
"grad_norm": 3.8858108520507812,
|
| 819 |
+
"lr": 2.021687061117951e-07
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"update": 102,
|
| 823 |
+
"ce": 1.0655539631843567,
|
| 824 |
+
"kl": 0.0007623753335792571,
|
| 825 |
+
"drift_frac": 0.039963989052921534,
|
| 826 |
+
"grad_norm": 4.007508277893066,
|
| 827 |
+
"lr": 2.0025951886443687e-07
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"update": 103,
|
| 831 |
+
"ce": 1.112197618931532,
|
| 832 |
+
"kl": 0.0007868932843848597,
|
| 833 |
+
"drift_frac": 0.03515250887721777,
|
| 834 |
+
"grad_norm": 5.913398742675781,
|
| 835 |
+
"lr": 1.9834273289114634e-07
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"update": 104,
|
| 839 |
+
"ce": 0.6632561162114143,
|
| 840 |
+
"kl": 0.0007180713437264785,
|
| 841 |
+
"drift_frac": 0.0335660083219409,
|
| 842 |
+
"grad_norm": 3.329941511154175,
|
| 843 |
+
"lr": 1.9641876127558867e-07
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"update": 105,
|
| 847 |
+
"ce": 0.42204682528972626,
|
| 848 |
+
"kl": 0.0004891064745606855,
|
| 849 |
+
"drift_frac": 0.022147178649902344,
|
| 850 |
+
"grad_norm": 2.448812484741211,
|
| 851 |
+
"lr": 1.9448801864999575e-07
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"update": 106,
|
| 855 |
+
"ce": 0.8198431134223938,
|
| 856 |
+
"kl": 0.0007205388610600494,
|
| 857 |
+
"drift_frac": 0.04275422589853406,
|
| 858 |
+
"grad_norm": 3.3530867099761963,
|
| 859 |
+
"lr": 1.9255092110581e-07
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"update": 107,
|
| 863 |
+
"ce": 0.7668614983558655,
|
| 864 |
+
"kl": 0.0004803506162716076,
|
| 865 |
+
"drift_frac": 0.018458958016708493,
|
| 866 |
+
"grad_norm": 3.58644437789917,
|
| 867 |
+
"lr": 1.9060788610401251e-07
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"update": 108,
|
| 871 |
+
"ce": 0.8220430240035057,
|
| 872 |
+
"kl": 0.0007829938549548388,
|
| 873 |
+
"drift_frac": 0.04378982027992606,
|
| 874 |
+
"grad_norm": 3.396223545074463,
|
| 875 |
+
"lr": 1.8865933238515722e-07
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"update": 109,
|
| 879 |
+
"ce": 0.4214981645345688,
|
| 880 |
+
"kl": 0.00035308169026393443,
|
| 881 |
+
"drift_frac": 0.018620133749209344,
|
| 882 |
+
"grad_norm": 2.473634719848633,
|
| 883 |
+
"lr": 1.8670567987912822e-07
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"update": 110,
|
| 887 |
+
"ce": 0.8305295407772064,
|
| 888 |
+
"kl": 0.0007212849159259349,
|
| 889 |
+
"drift_frac": 0.041692549362778664,
|
| 890 |
+
"grad_norm": 3.391541004180908,
|
| 891 |
+
"lr": 1.847473496146417e-07
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"update": 111,
|
| 895 |
+
"ce": 0.4409872516989708,
|
| 896 |
+
"kl": 0.0004845829898840748,
|
| 897 |
+
"drift_frac": 0.021965014515444636,
|
| 898 |
+
"grad_norm": 2.5255942344665527,
|
| 899 |
+
"lr": 1.8278476362851056e-07
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"update": 112,
|
| 903 |
+
"ce": 0.5992020480334759,
|
| 904 |
+
"kl": 0.0004972513415850699,
|
| 905 |
+
"drift_frac": 0.026829699985682964,
|
| 906 |
+
"grad_norm": 2.650059938430786,
|
| 907 |
+
"lr": 1.8081834487469175e-07
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"update": 113,
|
| 911 |
+
"ce": 0.5650180764496326,
|
| 912 |
+
"kl": 0.0005084919903310947,
|
| 913 |
+
"drift_frac": 0.024554014904424548,
|
| 914 |
+
"grad_norm": 3.6175878047943115,
|
| 915 |
+
"lr": 1.788485171331365e-07
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"update": 114,
|
| 919 |
+
"ce": 0.7393236756324768,
|
| 920 |
+
"kl": 0.000608368129178416,
|
| 921 |
+
"drift_frac": 0.03467158041894436,
|
| 922 |
+
"grad_norm": 3.3019821643829346,
|
| 923 |
+
"lr": 1.768757049184615e-07
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"update": 115,
|
| 927 |
+
"ce": 1.1030493676662445,
|
| 928 |
+
"kl": 0.0009180889610433951,
|
| 929 |
+
"drift_frac": 0.04753405600786209,
|
| 930 |
+
"grad_norm": 3.5900726318359375,
|
| 931 |
+
"lr": 1.7490033338846305e-07
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"update": 116,
|
| 935 |
+
"ce": 0.6274227648973465,
|
| 936 |
+
"kl": 0.0006103207560954615,
|
| 937 |
+
"drift_frac": 0.030978168826550245,
|
| 938 |
+
"grad_norm": 2.7972660064697266,
|
| 939 |
+
"lr": 1.729228282524915e-07
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"update": 117,
|
| 943 |
+
"ce": 0.4507666192948818,
|
| 944 |
+
"kl": 0.0005694737046724185,
|
| 945 |
+
"drift_frac": 0.02701697521843016,
|
| 946 |
+
"grad_norm": 2.624803304672241,
|
| 947 |
+
"lr": 1.7094361567970744e-07
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"update": 118,
|
| 951 |
+
"ce": 0.728460431098938,
|
| 952 |
+
"kl": 0.0008221390671678819,
|
| 953 |
+
"drift_frac": 0.030463837552815676,
|
| 954 |
+
"grad_norm": 3.120927572250366,
|
| 955 |
+
"lr": 1.6896312220723873e-07
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"update": 119,
|
| 959 |
+
"ce": 0.5561112090945244,
|
| 960 |
+
"kl": 0.0006870661964057945,
|
| 961 |
+
"drift_frac": 0.03471838450059295,
|
| 962 |
+
"grad_norm": 2.87046217918396,
|
| 963 |
+
"lr": 1.6698177464825775e-07
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"update": 120,
|
| 967 |
+
"ce": 0.5233822092413902,
|
| 968 |
+
"kl": 0.0005151640471012797,
|
| 969 |
+
"drift_frac": 0.026004251092672348,
|
| 970 |
+
"grad_norm": 2.5489630699157715,
|
| 971 |
+
"lr": 1.65e-07
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"update": 121,
|
| 975 |
+
"ce": 0.4638753980398178,
|
| 976 |
+
"kl": 0.0005271815170999616,
|
| 977 |
+
"drift_frac": 0.023199203889817,
|
| 978 |
+
"grad_norm": 2.6025567054748535,
|
| 979 |
+
"lr": 1.6301822535174227e-07
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"update": 122,
|
| 983 |
+
"ce": 0.4416641481220722,
|
| 984 |
+
"kl": 0.00039752620068611577,
|
| 985 |
+
"drift_frac": 0.020889636129140854,
|
| 986 |
+
"grad_norm": 2.560997486114502,
|
| 987 |
+
"lr": 1.6103687779276131e-07
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"update": 123,
|
| 991 |
+
"ce": 0.6657969318330288,
|
| 992 |
+
"kl": 0.0005187654787732754,
|
| 993 |
+
"drift_frac": 0.022642727708444,
|
| 994 |
+
"grad_norm": 3.006624937057495,
|
| 995 |
+
"lr": 1.5905638432029252e-07
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"update": 124,
|
| 999 |
+
"ce": 0.3786514326930046,
|
| 1000 |
+
"kl": 0.0004238432229612954,
|
| 1001 |
+
"drift_frac": 0.01974433334544301,
|
| 1002 |
+
"grad_norm": 2.154917001724243,
|
| 1003 |
+
"lr": 1.5707717174750854e-07
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"update": 125,
|
| 1007 |
+
"ce": 0.28607265651226044,
|
| 1008 |
+
"kl": 0.00039017735980451107,
|
| 1009 |
+
"drift_frac": 0.019554174272343516,
|
| 1010 |
+
"grad_norm": 1.9036833047866821,
|
| 1011 |
+
"lr": 1.55099666611537e-07
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"update": 126,
|
| 1015 |
+
"ce": 0.3243948966264725,
|
| 1016 |
+
"kl": 0.00038184469303814694,
|
| 1017 |
+
"drift_frac": 0.016987845534458756,
|
| 1018 |
+
"grad_norm": 2.7309937477111816,
|
| 1019 |
+
"lr": 1.5312429508153851e-07
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"update": 127,
|
| 1023 |
+
"ce": 0.6109912097454071,
|
| 1024 |
+
"kl": 0.0005933811189606786,
|
| 1025 |
+
"drift_frac": 0.03465761058032513,
|
| 1026 |
+
"grad_norm": 2.846731185913086,
|
| 1027 |
+
"lr": 1.5115148286686358e-07
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"update": 128,
|
| 1031 |
+
"ce": 0.44294311851263046,
|
| 1032 |
+
"kl": 0.0003965318464906886,
|
| 1033 |
+
"drift_frac": 0.020122107351198792,
|
| 1034 |
+
"grad_norm": 2.1913511753082275,
|
| 1035 |
+
"lr": 1.491816551253082e-07
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"update": 129,
|
| 1039 |
+
"ce": 0.6252846419811249,
|
| 1040 |
+
"kl": 0.0006558734603459015,
|
| 1041 |
+
"drift_frac": 0.02859257929958403,
|
| 1042 |
+
"grad_norm": 2.729797840118408,
|
| 1043 |
+
"lr": 1.4721523637148945e-07
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"update": 130,
|
| 1047 |
+
"ce": 0.44925640523433685,
|
| 1048 |
+
"kl": 0.0004932531301165,
|
| 1049 |
+
"drift_frac": 0.024147415766492486,
|
| 1050 |
+
"grad_norm": 1.9889088869094849,
|
| 1051 |
+
"lr": 1.452526503853583e-07
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"update": 131,
|
| 1055 |
+
"ce": 0.4228949248790741,
|
| 1056 |
+
"kl": 0.00043604255915852264,
|
| 1057 |
+
"drift_frac": 0.020532865077257156,
|
| 1058 |
+
"grad_norm": 2.1206538677215576,
|
| 1059 |
+
"lr": 1.432943201208718e-07
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"update": 132,
|
| 1063 |
+
"ce": 0.8460842296481133,
|
| 1064 |
+
"kl": 0.000615623124758713,
|
| 1065 |
+
"drift_frac": 0.027619050117209554,
|
| 1066 |
+
"grad_norm": 3.994272232055664,
|
| 1067 |
+
"lr": 1.4134066761484282e-07
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"update": 133,
|
| 1071 |
+
"ce": 0.4404878318309784,
|
| 1072 |
+
"kl": 0.0006261239250306971,
|
| 1073 |
+
"drift_frac": 0.027321083238348365,
|
| 1074 |
+
"grad_norm": 2.634896993637085,
|
| 1075 |
+
"lr": 1.3939211389598745e-07
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"update": 134,
|
| 1079 |
+
"ce": 0.776138249784708,
|
| 1080 |
+
"kl": 0.000673248476232402,
|
| 1081 |
+
"drift_frac": 0.03718887525610626,
|
| 1082 |
+
"grad_norm": 3.7520740032196045,
|
| 1083 |
+
"lr": 1.3744907889419002e-07
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"update": 135,
|
| 1087 |
+
"ce": 0.41094744950532913,
|
| 1088 |
+
"kl": 0.00041897271876223385,
|
| 1089 |
+
"drift_frac": 0.016334223095327616,
|
| 1090 |
+
"grad_norm": 2.2119436264038086,
|
| 1091 |
+
"lr": 1.3551198135000424e-07
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"update": 136,
|
| 1095 |
+
"ce": 0.4504081681370735,
|
| 1096 |
+
"kl": 0.0006172348948894069,
|
| 1097 |
+
"drift_frac": 0.019951058086007833,
|
| 1098 |
+
"grad_norm": 2.3867528438568115,
|
| 1099 |
+
"lr": 1.3358123872441138e-07
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"update": 137,
|
| 1103 |
+
"ce": 0.45100362598896027,
|
| 1104 |
+
"kl": 0.00047169711615424603,
|
| 1105 |
+
"drift_frac": 0.024274895898997784,
|
| 1106 |
+
"grad_norm": 2.353516101837158,
|
| 1107 |
+
"lr": 1.3165726710885367e-07
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"update": 138,
|
| 1111 |
+
"ce": 0.3370257057249546,
|
| 1112 |
+
"kl": 0.00048369279102189466,
|
| 1113 |
+
"drift_frac": 0.022571426816284657,
|
| 1114 |
+
"grad_norm": 2.0769810676574707,
|
| 1115 |
+
"lr": 1.2974048113556314e-07
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"update": 139,
|
| 1119 |
+
"ce": 0.8149940818548203,
|
| 1120 |
+
"kl": 0.0008968895563157275,
|
| 1121 |
+
"drift_frac": 0.04659335967153311,
|
| 1122 |
+
"grad_norm": 2.441587209701538,
|
| 1123 |
+
"lr": 1.2783129388820492e-07
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"update": 140,
|
| 1127 |
+
"ce": 0.93794334679842,
|
| 1128 |
+
"kl": 0.0006751314867869951,
|
| 1129 |
+
"drift_frac": 0.03543083043769002,
|
| 1130 |
+
"grad_norm": 3.1647419929504395,
|
| 1131 |
+
"lr": 1.259301168128541e-07
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"update": 141,
|
| 1135 |
+
"ce": 0.4858764410018921,
|
| 1136 |
+
"kl": 0.000668645923724398,
|
| 1137 |
+
"drift_frac": 0.031081425491720438,
|
| 1138 |
+
"grad_norm": 2.784503936767578,
|
| 1139 |
+
"lr": 1.2403735962932562e-07
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"update": 142,
|
| 1143 |
+
"ce": 0.38377512991428375,
|
| 1144 |
+
"kl": 0.00042201951873721555,
|
| 1145 |
+
"drift_frac": 0.020332837477326393,
|
| 1146 |
+
"grad_norm": 2.1698544025421143,
|
| 1147 |
+
"lr": 1.221534302428762e-07
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"update": 143,
|
| 1151 |
+
"ce": 0.341024212539196,
|
| 1152 |
+
"kl": 0.0003989593460573815,
|
| 1153 |
+
"drift_frac": 0.018100851913914084,
|
| 1154 |
+
"grad_norm": 1.861494541168213,
|
| 1155 |
+
"lr": 1.2027873465629725e-07
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"update": 144,
|
| 1159 |
+
"ce": 0.6110884621739388,
|
| 1160 |
+
"kl": 0.0006132917915238068,
|
| 1161 |
+
"drift_frac": 0.03481879294849932,
|
| 1162 |
+
"grad_norm": 2.5760772228240967,
|
| 1163 |
+
"lr": 1.1841367688241787e-07
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"update": 145,
|
| 1167 |
+
"ce": 0.5656596198678017,
|
| 1168 |
+
"kl": 0.0006215472312760539,
|
| 1169 |
+
"drift_frac": 0.024013089947402477,
|
| 1170 |
+
"grad_norm": 2.2479093074798584,
|
| 1171 |
+
"lr": 1.1655865885703667e-07
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"update": 146,
|
| 1175 |
+
"ce": 0.2913961037993431,
|
| 1176 |
+
"kl": 0.00042873337952187285,
|
| 1177 |
+
"drift_frac": 0.01757917506620288,
|
| 1178 |
+
"grad_norm": 1.6709855794906616,
|
| 1179 |
+
"lr": 1.1471408035230103e-07
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"update": 147,
|
| 1183 |
+
"ce": 0.3744642324745655,
|
| 1184 |
+
"kl": 0.0005368441925384104,
|
| 1185 |
+
"drift_frac": 0.021240175934508443,
|
| 1186 |
+
"grad_norm": 1.967044711112976,
|
| 1187 |
+
"lr": 1.1288033889055297e-07
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"update": 148,
|
| 1191 |
+
"ce": 0.6395589262247086,
|
| 1192 |
+
"kl": 0.001547366482554935,
|
| 1193 |
+
"drift_frac": 0.038705209735780954,
|
| 1194 |
+
"grad_norm": 2.375321865081787,
|
| 1195 |
+
"lr": 1.1105782965865981e-07
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"update": 149,
|
| 1199 |
+
"ce": 0.5652287751436234,
|
| 1200 |
+
"kl": 0.0004976389463990927,
|
| 1201 |
+
"drift_frac": 0.020294884452596307,
|
| 1202 |
+
"grad_norm": 2.2742717266082764,
|
| 1203 |
+
"lr": 1.0924694542284757e-07
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"update": 150,
|
| 1207 |
+
"ce": 0.7202664762735367,
|
| 1208 |
+
"kl": 0.000658013996144291,
|
| 1209 |
+
"drift_frac": 0.030844782944768667,
|
| 1210 |
+
"grad_norm": 2.1639113426208496,
|
| 1211 |
+
"lr": 1.0744807644405715e-07
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"update": 151,
|
| 1215 |
+
"ce": 0.3629290461540222,
|
| 1216 |
+
"kl": 0.0005562948063015938,
|
| 1217 |
+
"drift_frac": 0.018944226438179612,
|
| 1218 |
+
"grad_norm": 1.875972032546997,
|
| 1219 |
+
"lr": 1.056616103938393e-07
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"update": 152,
|
| 1223 |
+
"ce": 0.38130757957696915,
|
| 1224 |
+
"kl": 0.00047892545990180224,
|
| 1225 |
+
"drift_frac": 0.02467057667672634,
|
| 1226 |
+
"grad_norm": 2.1107304096221924,
|
| 1227 |
+
"lr": 1.0388793227080828e-07
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"update": 153,
|
| 1231 |
+
"ce": 1.7272626161575317,
|
| 1232 |
+
"kl": 0.0008666292123962194,
|
| 1233 |
+
"drift_frac": 0.03886930271983147,
|
| 1234 |
+
"grad_norm": 4.406638145446777,
|
| 1235 |
+
"lr": 1.0212742431767142e-07
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"update": 154,
|
| 1239 |
+
"ce": 0.5201831571757793,
|
| 1240 |
+
"kl": 0.0007942379961605184,
|
| 1241 |
+
"drift_frac": 0.027398436330258846,
|
| 1242 |
+
"grad_norm": 3.442302703857422,
|
| 1243 |
+
"lr": 1.0038046593885249e-07
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"update": 155,
|
| 1247 |
+
"ce": 1.3800606429576874,
|
| 1248 |
+
"kl": 0.0012993613781873137,
|
| 1249 |
+
"drift_frac": 0.05054049612954259,
|
| 1250 |
+
"grad_norm": 5.470850944519043,
|
| 1251 |
+
"lr": 9.864743361872684e-08
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"update": 156,
|
| 1255 |
+
"ce": 1.8901871144771576,
|
| 1256 |
+
"kl": 0.0018831207416951656,
|
| 1257 |
+
"drift_frac": 0.09361459594219923,
|
| 1258 |
+
"grad_norm": 11.897987365722656,
|
| 1259 |
+
"lr": 9.692870084048592e-08
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"update": 157,
|
| 1263 |
+
"ce": 1.4489086866378784,
|
| 1264 |
+
"kl": 0.002566061506513506,
|
| 1265 |
+
"drift_frac": 0.12067636102437973,
|
| 1266 |
+
"grad_norm": 10.22236442565918,
|
| 1267 |
+
"lr": 9.522463800564831e-08
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"update": 158,
|
| 1271 |
+
"ce": 1.6906611025333405,
|
| 1272 |
+
"kl": 0.005034955567680299,
|
| 1273 |
+
"drift_frac": 0.10002507548779249,
|
| 1274 |
+
"grad_norm": 13.01463508605957,
|
| 1275 |
+
"lr": 9.353561235423514e-08
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"update": 159,
|
| 1279 |
+
"ce": 1.2614936530590057,
|
| 1280 |
+
"kl": 0.0027290915022604167,
|
| 1281 |
+
"drift_frac": 0.10823673009872437,
|
| 1282 |
+
"grad_norm": 7.796757221221924,
|
| 1283 |
+
"lr": 9.186198788562676e-08
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"update": 160,
|
| 1287 |
+
"ce": 1.5873446464538574,
|
| 1288 |
+
"kl": 0.002485283272108063,
|
| 1289 |
+
"drift_frac": 0.09638062305748463,
|
| 1290 |
+
"grad_norm": 9.252289772033691,
|
| 1291 |
+
"lr": 9.020412528011771e-08
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"update": 161,
|
| 1295 |
+
"ce": 1.4573603421449661,
|
| 1296 |
+
"kl": 0.0016902399074751884,
|
| 1297 |
+
"drift_frac": 0.08514255098998547,
|
| 1298 |
+
"grad_norm": 9.871357917785645,
|
| 1299 |
+
"lr": 8.85623818211872e-08
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"update": 162,
|
| 1303 |
+
"ce": 1.3350442498922348,
|
| 1304 |
+
"kl": 0.003438036219449714,
|
| 1305 |
+
"drift_frac": 0.10395260900259018,
|
| 1306 |
+
"grad_norm": 12.417257308959961,
|
| 1307 |
+
"lr": 8.693711131850127e-08
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"update": 163,
|
| 1311 |
+
"ce": 1.507317915558815,
|
| 1312 |
+
"kl": 0.002689231128897518,
|
| 1313 |
+
"drift_frac": 0.11660680919885635,
|
| 1314 |
+
"grad_norm": 10.293481826782227,
|
| 1315 |
+
"lr": 8.532866403166397e-08
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"update": 164,
|
| 1319 |
+
"ce": 1.0273857712745667,
|
| 1320 |
+
"kl": 0.001990796605241485,
|
| 1321 |
+
"drift_frac": 0.08135003969073296,
|
| 1322 |
+
"grad_norm": 5.631631851196289,
|
| 1323 |
+
"lr": 8.37373865947333e-08
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"update": 165,
|
| 1327 |
+
"ce": 1.2940095663070679,
|
| 1328 |
+
"kl": 0.002247949334559962,
|
| 1329 |
+
"drift_frac": 0.07621539197862148,
|
| 1330 |
+
"grad_norm": 7.392284393310547,
|
| 1331 |
+
"lr": 8.216362194151885e-08
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"update": 166,
|
| 1335 |
+
"ce": 1.3492067605257034,
|
| 1336 |
+
"kl": 0.0020093094499316067,
|
| 1337 |
+
"drift_frac": 0.10303397104144096,
|
| 1338 |
+
"grad_norm": 7.118369102478027,
|
| 1339 |
+
"lr": 8.060770923167634e-08
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"update": 167,
|
| 1343 |
+
"ce": 1.530134528875351,
|
| 1344 |
+
"kl": 0.0026551474584266543,
|
| 1345 |
+
"drift_frac": 0.07573109958320856,
|
| 1346 |
+
"grad_norm": 7.868183135986328,
|
| 1347 |
+
"lr": 7.906998377761617e-08
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"update": 168,
|
| 1351 |
+
"ce": 1.1373808979988098,
|
| 1352 |
+
"kl": 0.0018624630174599588,
|
| 1353 |
+
"drift_frac": 0.06918956618756056,
|
| 1354 |
+
"grad_norm": 6.373600959777832,
|
| 1355 |
+
"lr": 7.755077697224036e-08
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"update": 169,
|
| 1359 |
+
"ce": 1.4417333900928497,
|
| 1360 |
+
"kl": 0.002085537591483444,
|
| 1361 |
+
"drift_frac": 0.06359903421252966,
|
| 1362 |
+
"grad_norm": 7.121034622192383,
|
| 1363 |
+
"lr": 7.605041621752503e-08
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"update": 170,
|
| 1367 |
+
"ce": 1.3691628575325012,
|
| 1368 |
+
"kl": 0.002684765961021185,
|
| 1369 |
+
"drift_frac": 0.08002958726137877,
|
| 1370 |
+
"grad_norm": 6.558362007141113,
|
| 1371 |
+
"lr": 7.4569224853962e-08
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"update": 171,
|
| 1375 |
+
"ce": 1.4715987592935562,
|
| 1376 |
+
"kl": 0.0024580707540735602,
|
| 1377 |
+
"drift_frac": 0.07233430724591017,
|
| 1378 |
+
"grad_norm": 6.935791492462158,
|
| 1379 |
+
"lr": 7.310752209087652e-08
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"update": 172,
|
| 1383 |
+
"ce": 0.8169498667120934,
|
| 1384 |
+
"kl": 0.0011616248957579955,
|
| 1385 |
+
"drift_frac": 0.05478035658597946,
|
| 1386 |
+
"grad_norm": 4.957376003265381,
|
| 1387 |
+
"lr": 7.166562293763488e-08
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"update": 173,
|
| 1391 |
+
"ce": 1.1998403072357178,
|
| 1392 |
+
"kl": 0.0018441294378135353,
|
| 1393 |
+
"drift_frac": 0.05741545557975769,
|
| 1394 |
+
"grad_norm": 7.758091926574707,
|
| 1395 |
+
"lr": 7.024383813575709e-08
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"update": 174,
|
| 1399 |
+
"ce": 1.5139262974262238,
|
| 1400 |
+
"kl": 0.0029597834800370038,
|
| 1401 |
+
"drift_frac": 0.07513836305588484,
|
| 1402 |
+
"grad_norm": 7.503878593444824,
|
| 1403 |
+
"lr": 6.884247409194954e-08
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"update": 175,
|
| 1407 |
+
"ce": 1.4761824309825897,
|
| 1408 |
+
"kl": 0.0017411693406756967,
|
| 1409 |
+
"drift_frac": 0.0861028116196394,
|
| 1410 |
+
"grad_norm": 8.098918914794922,
|
| 1411 |
+
"lr": 6.746183281207173e-08
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"update": 176,
|
| 1415 |
+
"ce": 1.008913516998291,
|
| 1416 |
+
"kl": 0.002131493834895082,
|
| 1417 |
+
"drift_frac": 0.06957422196865082,
|
| 1418 |
+
"grad_norm": 5.3707427978515625,
|
| 1419 |
+
"lr": 6.610221183605142e-08
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"update": 177,
|
| 1423 |
+
"ce": 1.214387446641922,
|
| 1424 |
+
"kl": 0.0019587668066378683,
|
| 1425 |
+
"drift_frac": 0.07177617214620113,
|
| 1426 |
+
"grad_norm": 5.382661819458008,
|
| 1427 |
+
"lr": 6.476390417376233e-08
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"update": 178,
|
| 1431 |
+
"ce": 1.102199137210846,
|
| 1432 |
+
"kl": 0.0015105211350601166,
|
| 1433 |
+
"drift_frac": 0.0737796388566494,
|
| 1434 |
+
"grad_norm": 4.928555965423584,
|
| 1435 |
+
"lr": 6.344719824187784e-08
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"update": 179,
|
| 1439 |
+
"ce": 1.1914029568433762,
|
| 1440 |
+
"kl": 0.0016002963529899716,
|
| 1441 |
+
"drift_frac": 0.06519563123583794,
|
| 1442 |
+
"grad_norm": 5.35493278503418,
|
| 1443 |
+
"lr": 6.215237780171503e-08
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"update": 180,
|
| 1447 |
+
"ce": 0.8325778022408485,
|
| 1448 |
+
"kl": 0.0010342434397898614,
|
| 1449 |
+
"drift_frac": 0.05343656614422798,
|
| 1450 |
+
"grad_norm": 5.0492048263549805,
|
| 1451 |
+
"lr": 6.087972189808151e-08
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"update": 181,
|
| 1455 |
+
"ce": 1.006698563694954,
|
| 1456 |
+
"kl": 0.0016810436791274697,
|
| 1457 |
+
"drift_frac": 0.07299915235489607,
|
| 1458 |
+
"grad_norm": 5.213586330413818,
|
| 1459 |
+
"lr": 5.962950479913879e-08
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"update": 182,
|
| 1463 |
+
"ce": 1.2083407193422318,
|
| 1464 |
+
"kl": 0.0018253272282890975,
|
| 1465 |
+
"drift_frac": 0.06829093582928181,
|
| 1466 |
+
"grad_norm": 5.045332431793213,
|
| 1467 |
+
"lr": 5.840199593729544e-08
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"update": 183,
|
| 1471 |
+
"ce": 1.521203488111496,
|
| 1472 |
+
"kl": 0.00185937769128941,
|
| 1473 |
+
"drift_frac": 0.07698439713567495,
|
| 1474 |
+
"grad_norm": 6.2342047691345215,
|
| 1475 |
+
"lr": 5.7197459851141894e-08
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"update": 184,
|
| 1479 |
+
"ce": 1.1030659079551697,
|
| 1480 |
+
"kl": 0.0013469437108142301,
|
| 1481 |
+
"drift_frac": 0.07039039954543114,
|
| 1482 |
+
"grad_norm": 5.275223731994629,
|
| 1483 |
+
"lr": 5.6016156128440205e-08
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"update": 185,
|
| 1487 |
+
"ce": 1.4310984015464783,
|
| 1488 |
+
"kl": 0.00256348037510179,
|
| 1489 |
+
"drift_frac": 0.07527569029480219,
|
| 1490 |
+
"grad_norm": 6.160255432128906,
|
| 1491 |
+
"lr": 5.485833935018078e-08
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"update": 186,
|
| 1495 |
+
"ce": 0.9513466507196426,
|
| 1496 |
+
"kl": 0.001210594069561921,
|
| 1497 |
+
"drift_frac": 0.06174378003925085,
|
| 1498 |
+
"grad_norm": 4.909823894500732,
|
| 1499 |
+
"lr": 5.372425903571783e-08
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"update": 187,
|
| 1503 |
+
"ce": 1.1369631737470627,
|
| 1504 |
+
"kl": 0.0012867886543972418,
|
| 1505 |
+
"drift_frac": 0.058397949673235416,
|
| 1506 |
+
"grad_norm": 4.9955525398254395,
|
| 1507 |
+
"lr": 5.261415958899596e-08
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"update": 188,
|
| 1511 |
+
"ce": 1.1622545272111893,
|
| 1512 |
+
"kl": 0.0019296630343887955,
|
| 1513 |
+
"drift_frac": 0.07178920134902,
|
| 1514 |
+
"grad_norm": 5.206904411315918,
|
| 1515 |
+
"lr": 5.152828024587896e-08
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"update": 189,
|
| 1519 |
+
"ce": 1.121278777718544,
|
| 1520 |
+
"kl": 0.0017273076518904418,
|
| 1521 |
+
"drift_frac": 0.05239789001643658,
|
| 1522 |
+
"grad_norm": 4.7204203605651855,
|
| 1523 |
+
"lr": 5.046685502259252e-08
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"update": 190,
|
| 1527 |
+
"ce": 1.6305065602064133,
|
| 1528 |
+
"kl": 0.002635970973642543,
|
| 1529 |
+
"drift_frac": 0.07841737661510706,
|
| 1530 |
+
"grad_norm": 6.883478164672852,
|
| 1531 |
+
"lr": 4.943011266529166e-08
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"update": 191,
|
| 1535 |
+
"ce": 1.0282090306282043,
|
| 1536 |
+
"kl": 0.001776694887666963,
|
| 1537 |
+
"drift_frac": 0.07646714430302382,
|
| 1538 |
+
"grad_norm": 4.443524360656738,
|
| 1539 |
+
"lr": 4.8418276600764126e-08
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"update": 192,
|
| 1543 |
+
"ce": 1.2561487555503845,
|
| 1544 |
+
"kl": 0.0024570801469963044,
|
| 1545 |
+
"drift_frac": 0.06944671645760536,
|
| 1546 |
+
"grad_norm": 5.183022975921631,
|
| 1547 |
+
"lr": 4.7431564888279896e-08
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"update": 193,
|
| 1551 |
+
"ce": 1.431408315896988,
|
| 1552 |
+
"kl": 0.0016825728525873274,
|
| 1553 |
+
"drift_frac": 0.06543282140046358,
|
| 1554 |
+
"grad_norm": 5.991611957550049,
|
| 1555 |
+
"lr": 4.647019017259753e-08
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"update": 194,
|
| 1559 |
+
"ce": 1.1902555599808693,
|
| 1560 |
+
"kl": 0.0012643930385820568,
|
| 1561 |
+
"drift_frac": 0.05330218095332384,
|
| 1562 |
+
"grad_norm": 5.109584808349609,
|
| 1563 |
+
"lr": 4.553435963813754e-08
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"update": 195,
|
| 1567 |
+
"ce": 1.1378628760576248,
|
| 1568 |
+
"kl": 0.0024382196716032922,
|
| 1569 |
+
"drift_frac": 0.06478629633784294,
|
| 1570 |
+
"grad_norm": 4.543624401092529,
|
| 1571 |
+
"lr": 4.4624274964332265e-08
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"update": 196,
|
| 1575 |
+
"ce": 1.293069213628769,
|
| 1576 |
+
"kl": 0.0018263478996232152,
|
| 1577 |
+
"drift_frac": 0.07024340331554413,
|
| 1578 |
+
"grad_norm": 5.261373519897461,
|
| 1579 |
+
"lr": 4.3740132282162393e-08
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"update": 197,
|
| 1583 |
+
"ce": 1.2165727615356445,
|
| 1584 |
+
"kl": 0.0012203293154016137,
|
| 1585 |
+
"drift_frac": 0.06269884761422873,
|
| 1586 |
+
"grad_norm": 4.533685684204102,
|
| 1587 |
+
"lr": 4.288212213188888e-08
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"update": 198,
|
| 1591 |
+
"ce": 1.1645057201385498,
|
| 1592 |
+
"kl": 0.0016932099679252133,
|
| 1593 |
+
"drift_frac": 0.0602649450302124,
|
| 1594 |
+
"grad_norm": 3.9267406463623047,
|
| 1595 |
+
"lr": 4.205042942199001e-08
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"update": 199,
|
| 1599 |
+
"ce": 1.2506205141544342,
|
| 1600 |
+
"kl": 0.001344346528640017,
|
| 1601 |
+
"drift_frac": 0.05963186081498861,
|
| 1602 |
+
"grad_norm": 5.498126983642578,
|
| 1603 |
+
"lr": 4.1245233389312107e-08
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"update": 200,
|
| 1607 |
+
"ce": 1.3444381207227707,
|
| 1608 |
+
"kl": 0.0017289415263803676,
|
| 1609 |
+
"drift_frac": 0.05229849787428975,
|
| 1610 |
+
"grad_norm": 4.228652000427246,
|
| 1611 |
+
"lr": 4.0466707560442304e-08
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"update": 201,
|
| 1615 |
+
"ce": 0.8781314045190811,
|
| 1616 |
+
"kl": 0.0013024866639170796,
|
| 1617 |
+
"drift_frac": 0.05180249735713005,
|
| 1618 |
+
"grad_norm": 3.4058103561401367,
|
| 1619 |
+
"lr": 3.971501971431228e-08
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"update": 202,
|
| 1623 |
+
"ce": 1.4038123190402985,
|
| 1624 |
+
"kl": 0.0017209083016496152,
|
| 1625 |
+
"drift_frac": 0.060743595473468304,
|
| 1626 |
+
"grad_norm": 5.841649532318115,
|
| 1627 |
+
"lr": 3.8990331846040294e-08
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"update": 203,
|
| 1631 |
+
"ce": 1.2735214531421661,
|
| 1632 |
+
"kl": 0.0015042233862914145,
|
| 1633 |
+
"drift_frac": 0.055322395637631416,
|
| 1634 |
+
"grad_norm": 4.8987884521484375,
|
| 1635 |
+
"lr": 3.829280013202e-08
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"update": 204,
|
| 1639 |
+
"ce": 1.280187875032425,
|
| 1640 |
+
"kl": 0.0017483088304288685,
|
| 1641 |
+
"drift_frac": 0.06469053961336613,
|
| 1642 |
+
"grad_norm": 5.366652965545654,
|
| 1643 |
+
"lr": 3.762257489626299e-08
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"update": 205,
|
| 1647 |
+
"ce": 1.2743341028690338,
|
| 1648 |
+
"kl": 0.0014409445284400135,
|
| 1649 |
+
"drift_frac": 0.0629212036728859,
|
| 1650 |
+
"grad_norm": 4.312648773193359,
|
| 1651 |
+
"lr": 3.697980057800283e-08
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"update": 206,
|
| 1655 |
+
"ce": 1.2756515741348267,
|
| 1656 |
+
"kl": 0.0015334409836214036,
|
| 1657 |
+
"drift_frac": 0.05982220824807882,
|
| 1658 |
+
"grad_norm": 3.587595224380493,
|
| 1659 |
+
"lr": 3.636461570056706e-08
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"update": 207,
|
| 1663 |
+
"ce": 0.8883579075336456,
|
| 1664 |
+
"kl": 0.0013827572402078658,
|
| 1665 |
+
"drift_frac": 0.053972901310771704,
|
| 1666 |
+
"grad_norm": 3.2399964332580566,
|
| 1667 |
+
"lr": 3.5777152841524294e-08
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"update": 208,
|
| 1671 |
+
"ce": 0.6809385791420937,
|
| 1672 |
+
"kl": 0.0008741230340092443,
|
| 1673 |
+
"drift_frac": 0.03655677242204547,
|
| 1674 |
+
"grad_norm": 3.0585238933563232,
|
| 1675 |
+
"lr": 3.5217538604112596e-08
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"update": 209,
|
| 1679 |
+
"ce": 1.1473584175109863,
|
| 1680 |
+
"kl": 0.0010280547721777111,
|
| 1681 |
+
"drift_frac": 0.04543887358158827,
|
| 1682 |
+
"grad_norm": 3.9134297370910645,
|
| 1683 |
+
"lr": 3.4685893589955436e-08
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"update": 210,
|
| 1687 |
+
"ce": 1.3511615619063377,
|
| 1688 |
+
"kl": 0.0011198662978131324,
|
| 1689 |
+
"drift_frac": 0.051833643577992916,
|
| 1690 |
+
"grad_norm": 3.888165235519409,
|
| 1691 |
+
"lr": 3.4182332373071024e-08
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"update": 211,
|
| 1695 |
+
"ce": 0.683077335357666,
|
| 1696 |
+
"kl": 0.0007411860206048004,
|
| 1697 |
+
"drift_frac": 0.03223559819161892,
|
| 1698 |
+
"grad_norm": 3.1728744506835938,
|
| 1699 |
+
"lr": 3.370696347518056e-08
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"update": 212,
|
| 1703 |
+
"ce": 0.9470455944538116,
|
| 1704 |
+
"kl": 0.0009859194760792889,
|
| 1705 |
+
"drift_frac": 0.0481146015226841,
|
| 1706 |
+
"grad_norm": 3.1459767818450928,
|
| 1707 |
+
"lr": 3.325988934232101e-08
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"update": 213,
|
| 1711 |
+
"ce": 0.8747119605541229,
|
| 1712 |
+
"kl": 0.0008092402567854151,
|
| 1713 |
+
"drift_frac": 0.040138402953743935,
|
| 1714 |
+
"grad_norm": 3.4226956367492676,
|
| 1715 |
+
"lr": 3.284120632276702e-08
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"update": 214,
|
| 1719 |
+
"ce": 0.5474836006760597,
|
| 1720 |
+
"kl": 0.000918326317332685,
|
| 1721 |
+
"drift_frac": 0.03179580485448241,
|
| 1722 |
+
"grad_norm": 2.7624382972717285,
|
| 1723 |
+
"lr": 3.245100464626704e-08
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"update": 215,
|
| 1727 |
+
"ce": 0.7664187401533127,
|
| 1728 |
+
"kl": 0.001246710744453594,
|
| 1729 |
+
"drift_frac": 0.04220534581691027,
|
| 1730 |
+
"grad_norm": 3.793031930923462,
|
| 1731 |
+
"lr": 3.208936840459808e-08
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"update": 216,
|
| 1735 |
+
"ce": 0.9004264697432518,
|
| 1736 |
+
"kl": 0.0010518897033762187,
|
| 1737 |
+
"drift_frac": 0.044356548227369785,
|
| 1738 |
+
"grad_norm": 3.09504771232605,
|
| 1739 |
+
"lr": 3.1756375533443225e-08
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"update": 217,
|
| 1743 |
+
"ce": 0.6116755679249763,
|
| 1744 |
+
"kl": 0.0007651011983398348,
|
| 1745 |
+
"drift_frac": 0.0390096134506166,
|
| 1746 |
+
"grad_norm": 2.9167044162750244,
|
| 1747 |
+
"lr": 3.1452097795595797e-08
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"update": 218,
|
| 1751 |
+
"ce": 0.9062808305025101,
|
| 1752 |
+
"kl": 0.000979409072897397,
|
| 1753 |
+
"drift_frac": 0.04121864354237914,
|
| 1754 |
+
"grad_norm": 2.6760482788085938,
|
| 1755 |
+
"lr": 3.117660076549389e-08
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"update": 219,
|
| 1759 |
+
"ce": 0.7792034074664116,
|
| 1760 |
+
"kl": 0.0008861165551934391,
|
| 1761 |
+
"drift_frac": 0.04058573255315423,
|
| 1762 |
+
"grad_norm": 2.669189929962158,
|
| 1763 |
+
"lr": 3.0929943815088554e-08
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"update": 220,
|
| 1767 |
+
"ce": 1.4320838451385498,
|
| 1768 |
+
"kl": 0.001257634983630851,
|
| 1769 |
+
"drift_frac": 0.0643214276060462,
|
| 1770 |
+
"grad_norm": 3.561976194381714,
|
| 1771 |
+
"lr": 3.071218010104854e-08
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"update": 221,
|
| 1775 |
+
"ce": 0.4843621179461479,
|
| 1776 |
+
"kl": 0.0007362240139627829,
|
| 1777 |
+
"drift_frac": 0.029081282205879688,
|
| 1778 |
+
"grad_norm": 2.3666064739227295,
|
| 1779 |
+
"lr": 3.052335655330466e-08
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"update": 222,
|
| 1783 |
+
"ce": 0.5831042677164078,
|
| 1784 |
+
"kl": 0.0006797644891776145,
|
| 1785 |
+
"drift_frac": 0.03326878510415554,
|
| 1786 |
+
"grad_norm": 2.219430446624756,
|
| 1787 |
+
"lr": 3.0363513864935994e-08
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"update": 223,
|
| 1791 |
+
"ce": 0.4824071228504181,
|
| 1792 |
+
"kl": 0.0005861124082002789,
|
| 1793 |
+
"drift_frac": 0.02908932324498892,
|
| 1794 |
+
"grad_norm": 2.1191742420196533,
|
| 1795 |
+
"lr": 3.0232686483400124e-08
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"update": 224,
|
| 1799 |
+
"ce": 1.0188091099262238,
|
| 1800 |
+
"kl": 0.001389817611197941,
|
| 1801 |
+
"drift_frac": 0.04455316998064518,
|
| 1802 |
+
"grad_norm": 3.8705451488494873,
|
| 1803 |
+
"lr": 3.0130902603109497e-08
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"update": 225,
|
| 1807 |
+
"ce": 0.47116606682538986,
|
| 1808 |
+
"kl": 0.0005643544427584857,
|
| 1809 |
+
"drift_frac": 0.02538590133190155,
|
| 1810 |
+
"grad_norm": 2.0716733932495117,
|
| 1811 |
+
"lr": 3.0058184159355216e-08
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"update": 226,
|
| 1815 |
+
"ce": 0.48285093903541565,
|
| 1816 |
+
"kl": 0.0005785107059637085,
|
| 1817 |
+
"drift_frac": 0.026091191917657852,
|
| 1818 |
+
"grad_norm": 2.4227421283721924,
|
| 1819 |
+
"lr": 3.0014546823579825e-08
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"update": 227,
|
| 1823 |
+
"ce": 1.187558501958847,
|
| 1824 |
+
"kl": 0.001187696383567527,
|
| 1825 |
+
"drift_frac": 0.05528997164219618,
|
| 1826 |
+
"grad_norm": 3.054713487625122,
|
| 1827 |
+
"lr": 3e-08
|
| 1828 |
+
}
|
| 1829 |
+
],
|
| 1830 |
+
"source_path": "/marimo/budgie_information_rl_v1/v3_reasoning_sft_source.pt",
|
| 1831 |
+
"parameter_count_change": 0,
|
| 1832 |
+
"benchmark_family_rows": 0
|
| 1833 |
+
}
|