Llama-2-70b-chat-hf | RC_project Overview – Weights & Biases

Skip to main content

Nonenman's group workspace

Group: Llama-2-70b-chat-hf

1-20

of 85

Tags

Notes

If the question is unanswerable, say #unanswerable#

Author

State

Finished

Start time

March 28th, 2024 4:06:56 AM

Runtime

1h 14m 36s

Tracked hours

1h 14m 33s

Run path

nonenman/RC_project/160cl9mg

OS

Linux-4.18.0-372.96.1.el8_6.x86_64-x86_64-with-glibc2.28

Python version

3.11.7

Command

/pfs/data5/home/as/as_as/as_nonenman/python_scripts5/zero-shot_prompt_refrain_from_answering.py

System Hardware

CPU count	64
Logical CPU count	128
GPU count	1
GPU type	NVIDIA H100 PCIe

W&B CLI Version

0.16.2

Group

Llama-2-70b-chat-hf

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 8 keys
- data:
  "test"
- do_sample:
  true
- max_new_tokens:
  512
- num_beams:
  1
- prompt:
  "zero-shot"
- prompt_format:
  "Passage: #{context}# Question: #{question}#"
- temperature:
  0.1
- top_p:
  0.95

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 12 keys
- eval/exact_match:
  30.727023319615913
- eval/f1:
  54.80441141740432
- eval/nr_unanswered_questions:
  42
- eval/precision:
  51.58179297483369
- eval/recall:
  71.49559660556247
- model/hidden_size:
  8,192
- model/max_position_embeddings:
  4,096
- model/num_attention_heads:
  64
- model/num_hidden_layers:
  80
- model/vocab_size:
  32,000
- train/counter:
  1,500
- train/total_runtime:
  70.32009898053366

This run produced these artifacts as outputs. Total: 1. Learn more

run-160cl9mg-history:v0

Loading...