Views
No views yet
2060) from run qwen35_2b_lora_8_small_lr_short_form_prompt.Qwen/Qwen3.5-2B on AmbigQA.step=1620, epoch=3).sewon/ambig_qa (config_name=light)6e667596df70f17ba3c8e7be4b7361f6be8b60f8train, validation=validationdev_fraction=0.1, split_seed=1729FalseYou answer open-domain factoid questions with a short-form answer only.
Reply with just the answer string: typically a name, place, organization, title, date, number, or brief factual phrase (often about 1–4 words). Do not write a sentence, explanation, or preamble.
Examples of the expected style:
- Who painted The Persistence of Memory? → Salvador Dalí
- When was the first iPhone released in the United States? → June 29, 2007
- What is the capital of New Zealand? → Wellington
- Who wrote Pride and Prejudice? → Jane Austen
- What is the chemical symbol for gold? → Au
- Who directed the 1994 film Pulp Fiction? → Quentin Tarantino
Formatting rules:
- Output the answer alone on a single line.
- Do not restate the question, add reasoning, markdown, bullets, or quotation marks around the answer.
- Prefer a canonical short form over a descriptive sentence.
- Give one answer only; do not list alternatives or hedge.
- Use natural short forms for dates and names (e.g. June 29, 2007; Jane Austen).resolved_config.yaml):1model:
2 name: Qwen/Qwen3.5-2B
3 revision: main
4 dtype: bfloat16
5 enable_thinking: false
6dataset:
7 name: sewon/ambig_qa
8 config_name: light
9 revision: 6e667596df70f17ba3c8e7be4b7361f6be8b60f8
10 train_split: train
11 validation_split: validation
12 dev_fraction: 0.1
13 split_seed: 1729
14 bootstrap: false
15 max_train_examples: null
16 max_dev_examples: null
17 system_prompt_set: true
18adaptation:
19 method: lora
20 lora:
21 r: 8
22 alpha: 16
23 dropout: 0.0
24 target_modules: all-linear
25 bias: none
26training:
27 num_train_epochs: 3.0
28 max_steps: -1
29 learning_rate: 5.0e-05
30 weight_decay: 0.01
31 warmup_ratio: 0.03
32 per_device_train_batch_size: 16
33 per_device_eval_batch_size: 32
34 gradient_accumulation_steps: 2
35 max_seq_length: 512
36 logging_steps: 10
37 save_every_steps: 150
38 save_total_limit: 3
39 gradient_checkpointing: false
40 bf16: true
41 tf32: true
42 max_grad_norm: 1.0
43 dataloader_num_workers: 2
44 adam_beta1: 0.9
45 adam_beta2: 0.999
46 adam_epsilon: 1.0e-08
47ensemble:
48 size: 5
49 base_seed: 42
50 this_member_index: 2
51 this_member_seed: 2060
52evaluation_subsets:
53 ambigqa:
54 subset_size: 128
55 seed: 1001
56 ifeval:
57 subset_size: 64
58 seed: 1002
59 mmlu:
60 subset_size: 228
61 seed: 1004| member | seed | steps | epochs | train_loss | AmbigQA acc | AmbigQA AlignScore | IFEval strict | MMLU acc |
|---|---|---|---|---|---|---|---|---|
| 0 | 42 | 1620 | 3 | 1.5179 | 0.0703 | 0.1238 | 0.5469 | 0.5833 |
| 1 | 1051 | 1620 | 3 | 1.5207 | 0.0781 | 0.1265 | 0.4531 | 0.6096 |
| 2 (this repo) | 2060 | 1620 | 3 | 1.5133 | 0.0703 | 0.1341 | 0.5781 | 0.6184 |
| 3 | 3069 | 1620 | 3 | 1.5196 | 0.0859 | 0.1321 | 0.5156 | 0.5877 |
| 4 | 4078 | 1620 | 3 | 1.5245 | 0.0703 | 0.1225 | 0.5312 | 0.5789 |
members/member_002/final/resolved_config.yaml — full resolved training configensemble_metrics.png — train/eval curves for the whole ensemble