language:
- en
license: apache-2.0
tags:
- text-generation-inference
datasets:
- HuggingFaceH4/ultrachat_200k
- openchat/openchat_sharegpt4_dataset
- Open-Orca/SlimOrca
inference: false
model-index:
- name: falcon-rw-1b-chat
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 35.58
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericzzz/falcon-rw-1b-chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 61.12
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericzzz/falcon-rw-1b-chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 24.51
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericzzz/falcon-rw-1b-chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 39.62
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericzzz/falcon-rw-1b-chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 61.72
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericzzz/falcon-rw-1b-chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 1.67
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericzzz/falcon-rw-1b-chat
name: Open LLM Leaderboard
pipeline_tag: text-generation
🌟 Falcon-RW-1B-Chat
Falcon-RW-1B-Chat is a conversational model with 1 billion parameters. It's a further refinement of the Falcon-RW-1B-Instruct-OpenOrca, trained on selected data from the HuggingFaceH4/ultrachat_200k and openchat/openchat_sharegpt4_dataset datasets.
✨Try it out at our Tiny Chat space running on free-tier hardware!✨
The underlying Falcon-RW-1B-Instruct-OpenOrca model is built on the Falcon-RW-1B, a causal decoder-only model. It has been instruction-finetuned using the Open-Orca/SlimOrca dataset.
🎯 Purpose
The Falcon-RW-1B-Chat aims to add conversational capabilities to the Falcon-RW-1B-Instruct-OpenOrca model. This initiative is driven by the need for a smaller, open-source, instruction-finetuned, ready-to-use model, suitable for users with limited computational resources, like lower-end consumer GPUs.
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 37.37 |
| AI2 Reasoning Challenge (25-Shot) | 35.58 |
| HellaSwag (10-Shot) | 61.12 |
| MMLU (5-Shot) | 24.51 |
| TruthfulQA (0-shot) | 39.62 |
| Winogrande (5-shot) | 61.72 |
| GSM8k (5-shot) | 1.67 |
📖 Example Code
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
model_name = "ericzzz/falcon-rw-1b-chat"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name, device_map="auto", torch_dtype=torch.bfloat16
)
chat_history = [
{"role": "user", "content": "Hello!"},
{"role": "assistant", "content": "Hello! How can I assist you today?"},
{"role": "user", "content": "Explain what AI is."},
]
input_ids = tokenizer.apply_chat_template(
chat_history, tokenize=True, add_generation_prompt=True, return_tensors="pt"
).to(model.device)
output_tokens = model.generate(
input_ids,
do_sample=True,
temperature=0.7,
repetition_penalty=1.05,
max_new_tokens=200,
)
output_text = tokenizer.decode(
output_tokens[0][len(input_ids[0]) :], skip_special_tokens=True
)
print(output_text)
⚠️ Limitations
This model may generate inaccurate or misleading information and is prone to hallucination, creating plausible but false narratives. It lacks the ability to discern factual content from fiction and may inadvertently produce biased, harmful or offensive content. Its understanding of complex, nuanced queries is limited. Users should be aware of this and verify any information obtained from the model.
The model is provided 'as is' without any warranties, and the creators are not liable for any damages arising from its use. Users are responsible for their interactions with the model.
📬 Contact
For further inquiries or feedback, please contact at [email protected].