Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
configs:
|
| 3 |
+
- config_name: rat
|
| 4 |
+
data_files:
|
| 5 |
+
- split: train
|
| 6 |
+
path: data/rat.jsonl
|
| 7 |
+
---
|
| 8 |
+
<!-- config:rat -->
|
| 9 |
+
## rat
|
| 10 |
+
|
| 11 |
+
4240 rows (4240 successful, 0 failed). finish_reason counts: {'stop': 4240}.
|
| 12 |
+
Model: Qwen/Qwen3-8B (revision b968826d9c46dd6066d109eabc6255188de91218), thinking on.
|
| 13 |
+
|
| 14 |
+
### Sampling (as sent to the server)
|
| 15 |
+
```json
|
| 16 |
+
{
|
| 17 |
+
"temperature": 0.9,
|
| 18 |
+
"top_p": 0.95,
|
| 19 |
+
"top_k": 20,
|
| 20 |
+
"min_p": 0.0,
|
| 21 |
+
"max_tokens": 32768,
|
| 22 |
+
"seed": "not sent"
|
| 23 |
+
}
|
| 24 |
+
```
|
| 25 |
+
|
| 26 |
+
### Runs
|
| 27 |
+
git sha(s): ae98fd5044474596e571b2f0f8bd8723e3f57335;
|
| 28 |
+
slurm job(s): 19358600.
|
| 29 |
+
Full run manifest: manifests/rat.json.
|
| 30 |
+
|
| 31 |
+
### Dataset provenance
|
| 32 |
+
```json
|
| 33 |
+
{
|
| 34 |
+
"dataset": "rat",
|
| 35 |
+
"rows_total": 212,
|
| 36 |
+
"instruction": "Respond with only the single word answer, nothing else.",
|
| 37 |
+
"source": {
|
| 38 |
+
"xlsx": "remote_association_test.xlsx",
|
| 39 |
+
"xlsx_sha256": "24e6ec274e6901294f82182878fa87f2c5a1b8a5c8948c5d769fb30c2a3bb36d",
|
| 40 |
+
"solutions_module": "remote_associations_test.py",
|
| 41 |
+
"solutions_module_sha256": "904562f242f46962eb6a21c8c922a698e12b4d3708d6989a73bdb8791da4aade",
|
| 42 |
+
"solution_entries_in_module": 312
|
| 43 |
+
},
|
| 44 |
+
"duplicate_cue_sets": [
|
| 45 |
+
[
|
| 46 |
+
"birthday",
|
| 47 |
+
"light",
|
| 48 |
+
"stick"
|
| 49 |
+
]
|
| 50 |
+
]
|
| 51 |
+
}
|
| 52 |
+
```
|
| 53 |
+
|
| 54 |
+
### Columns
|
| 55 |
+
| Column | Description |
|
| 56 |
+
|---|---|
|
| 57 |
+
| dataset | dataset name (also the config name of this subset) |
|
| 58 |
+
| instance_id | instance id within the dataset |
|
| 59 |
+
| sample_idx | index of this sample for the instance |
|
| 60 |
+
| prompt_sha1 | sha1 of the exact prompt string sent to the server |
|
| 61 |
+
| prompt_tokens | prompt length in tokens |
|
| 62 |
+
| content | raw model output, untouched (thinking and </think> inline) |
|
| 63 |
+
| finish_reason | vLLM finish_reason: stop or length |
|
| 64 |
+
| usage | vLLM usage block: prompt_tokens, completion_tokens, total_tokens |
|
| 65 |
+
| success | true if the request returned a generation |
|
| 66 |
+
| error | error string for a failed request, else null |
|
| 67 |
+
| attempts | requests made for this row (1 + retries) |
|
| 68 |
+
| latency_s | wall seconds for this request, from client send to response |
|
| 69 |
+
| ts | unix time the row was written |
|
| 70 |
+
|
| 71 |
+
<!-- /config:rat -->
|