acnagle commited on
Commit
73cf049
·
verified ·
1 Parent(s): 9e2405b

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +4 -4
README.md CHANGED
@@ -66,11 +66,11 @@ dataset_info:
66
 
67
  # Terminator-Qwen3-8B Training Data
68
 
69
- [Project page](https://terminator-llm.github.io/) | [Paper](https://arxiv.org/abs/2603.12529) | [Code](https://github.com/acnagle/terminator-llm) | [Trained weights](https://huggingface.co/acnagle/Terminator-Qwen3-8B)
70
 
71
  The tokenized, labeled data used to train [Terminator-Qwen3-8B](https://huggingface.co/acnagle/Terminator-Qwen3-8B). [Terminator](https://terminator-llm.github.io/) is a small network that predicts when a reasoning model has reached its final answer so the remaining reasoning can be skipped.
72
 
73
- Each row is one chain of thought (CoT) from Qwen3-8B on a math, coding or science problem, with per-token labels marking where the CoT first reaches its final answer. This dataset holds training data only. To evaluate on the test benchmarks, generate their CoTs with the [code](https://github.com/acnagle/terminator-llm).
74
 
75
  ## Contents
76
 
@@ -95,7 +95,7 @@ tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-8B")
95
  text = tokenizer.decode(row["input_ids"]) # the prompt and the CoT
96
  ```
97
 
98
- To train with the [code](https://github.com/acnagle/terminator-llm), download the files to the path its scripts expect, then run the training script from the repo root:
99
 
100
  ```bash
101
  hf download acnagle/Terminator-Qwen3-8B-Training-Data train.jsonl val.jsonl \
@@ -123,7 +123,7 @@ Each problem has 3 CoTs, sampled from Qwen3-8B with temperature 0.6, top-p 0.95,
123
 
124
  `prepare_training_data.py` tokenized those CoTs and built the labels with its default settings, holding out 50 problems per source for validation using seed 1337. The rows are grouped by source, so shuffle them before training.
125
 
126
- Rebuilding this data from scratch requires generating CoTs and rerunning the pipeline in the [code](https://github.com/acnagle/terminator-llm).
127
 
128
  ## Sources and licenses
129
 
 
66
 
67
  # Terminator-Qwen3-8B Training Data
68
 
69
+ [Project page](https://terminator-llm.github.io/) | [Paper](https://arxiv.org/abs/2603.12529) | [Code](https://github.com/acnagle/terminator) | [Trained weights](https://huggingface.co/acnagle/Terminator-Qwen3-8B)
70
 
71
  The tokenized, labeled data used to train [Terminator-Qwen3-8B](https://huggingface.co/acnagle/Terminator-Qwen3-8B). [Terminator](https://terminator-llm.github.io/) is a small network that predicts when a reasoning model has reached its final answer so the remaining reasoning can be skipped.
72
 
73
+ Each row is one chain of thought (CoT) from Qwen3-8B on a math, coding or science problem, with per-token labels marking where the CoT first reaches its final answer. This dataset holds training data only. To evaluate on the test benchmarks, generate their CoTs with the [code](https://github.com/acnagle/terminator).
74
 
75
  ## Contents
76
 
 
95
  text = tokenizer.decode(row["input_ids"]) # the prompt and the CoT
96
  ```
97
 
98
+ To train with the [code](https://github.com/acnagle/terminator), download the files to the path its scripts expect, then run the training script from the repo root:
99
 
100
  ```bash
101
  hf download acnagle/Terminator-Qwen3-8B-Training-Data train.jsonl val.jsonl \
 
123
 
124
  `prepare_training_data.py` tokenized those CoTs and built the labels with its default settings, holding out 50 problems per source for validation using seed 1337. The rows are grouped by source, so shuffle them before training.
125
 
126
+ Rebuilding this data from scratch requires generating CoTs and rerunning the pipeline in the [code](https://github.com/acnagle/terminator).
127
 
128
  ## Sources and licenses
129