wadhma commited on
Commit
f863f19
·
verified ·
1 Parent(s): 84980b2

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +59 -0
README.md CHANGED
@@ -1,5 +1,9 @@
1
  ---
2
  configs:
 
 
 
 
3
  - config_name: rat
4
  data_files:
5
  - split: train
@@ -128,3 +132,58 @@ Full run manifest: manifests/simpleqa_verified.json.
128
  | ts | unix time the row was written |
129
 
130
  <!-- /config:simpleqa_verified -->
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  configs:
3
+ - config_name: gsm8k
4
+ data_files:
5
+ - split: train
6
+ path: data/gsm8k.jsonl
7
  - config_name: rat
8
  data_files:
9
  - split: train
 
132
  | ts | unix time the row was written |
133
 
134
  <!-- /config:simpleqa_verified -->
135
+
136
+ <!-- config:gsm8k -->
137
+ ## gsm8k
138
+
139
+ 26380 rows (26380 successful, 0 failed). finish_reason counts: {'stop': 26380}.
140
+ Model: Qwen/Qwen3-8B (revision b968826d9c46dd6066d109eabc6255188de91218), thinking on.
141
+
142
+ ### Sampling (as sent to the server)
143
+ ```json
144
+ {
145
+ "temperature": 0.9,
146
+ "top_p": 0.95,
147
+ "top_k": 20,
148
+ "min_p": 0.0,
149
+ "max_tokens": 32768,
150
+ "seed": "not sent"
151
+ }
152
+ ```
153
+
154
+ ### Runs
155
+ git sha(s): 0a093ace61cdaba95c3476c347ab8d9ae1745ef4;
156
+ slurm job(s): 21164277.
157
+ Full run manifest: manifests/gsm8k.json.
158
+
159
+ ### Dataset provenance
160
+ ```json
161
+ {
162
+ "dataset": "gsm8k",
163
+ "repo": "openai/gsm8k",
164
+ "file": "gsm8k_test.jsonl",
165
+ "revision": "740312add88f781978c0658806c59bc2815b9866",
166
+ "sha256": "752adc99f23132275139f1f7be57126b08735a55daf0c56859ee6df70abafbac",
167
+ "rows_total": 1319,
168
+ "instruction": null
169
+ }
170
+ ```
171
+
172
+ ### Columns
173
+ | Column | Description |
174
+ |---|---|
175
+ | dataset | dataset name (also the config name of this subset) |
176
+ | instance_id | instance id within the dataset |
177
+ | sample_idx | index of this sample for the instance |
178
+ | prompt_sha1 | sha1 of the exact prompt string sent to the server |
179
+ | prompt_tokens | prompt length in tokens |
180
+ | content | raw model output, untouched (thinking and </think> inline) |
181
+ | finish_reason | vLLM finish_reason: stop or length |
182
+ | usage | vLLM usage block: prompt_tokens, completion_tokens, total_tokens |
183
+ | success | true if the request returned a generation |
184
+ | error | error string for a failed request, else null |
185
+ | attempts | requests made for this row (1 + retries) |
186
+ | latency_s | wall seconds for this request, from client send to response |
187
+ | ts | unix time the row was written |
188
+
189
+ <!-- /config:gsm8k -->