Instructions to use maxint-inc/Finsight-Qwen3-14B-CPT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use maxint-inc/Finsight-Qwen3-14B-CPT with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="maxint-inc/Finsight-Qwen3-14B-CPT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("maxint-inc/Finsight-Qwen3-14B-CPT")
model = AutoModelForCausalLM.from_pretrained("maxint-inc/Finsight-Qwen3-14B-CPT", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use maxint-inc/Finsight-Qwen3-14B-CPT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "maxint-inc/Finsight-Qwen3-14B-CPT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "maxint-inc/Finsight-Qwen3-14B-CPT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/maxint-inc/Finsight-Qwen3-14B-CPT

SGLang

How to use maxint-inc/Finsight-Qwen3-14B-CPT with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "maxint-inc/Finsight-Qwen3-14B-CPT" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "maxint-inc/Finsight-Qwen3-14B-CPT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "maxint-inc/Finsight-Qwen3-14B-CPT" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "maxint-inc/Finsight-Qwen3-14B-CPT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use maxint-inc/Finsight-Qwen3-14B-CPT with Docker Model Runner:
```
docker model run hf.co/maxint-inc/Finsight-Qwen3-14B-CPT
```

Finsight — Qwen3-14B Finance (CPT only)

This is the intermediate checkpoint from the Finsight finance-adaptation of Qwen/Qwen3-14B-Base: the model after continued pre-training on raw finance text, but before instruction tuning.

👉 For the finished, instruction-following model, use maxint-inc/Finsight-Qwen3-14B-CPT-SFT. This CPT-only repo is shared for research/ablation.

⚠️ Base-style checkpoint — it has no chat template and is not instruction-tuned. It completes text; it does not follow instructions well yet.

What this stage did

We continued pre-training the base model on ~330M tokens of raw finance text so it would better understand financial language before any instruction tuning:

Source	What it is
SEC 10-K filings (`anonymous-md/EDGAR_FILINGS_DATASET`)	Full annual-report text
Financial news (`ashraq/financial-news-articles`)	Market/company news
Industry finance corpus (`BAAI/IndustryCorpus_Finance`)	Long-form finance text

Trained with QLoRA (4-bit NF4 + LoRA r=64, α=128), 800 steps (~0.2B tokens), then merged to 16-bit. Corpus decontaminated against the benchmark test sets.

Results (on the FinBen `flare_*` finance benchmarks)

Same samples, greedy decoding, via lm-evaluation-harness.

Benchmark (skill)	Base	CPT (this)	Final (CPT→SFT)
FinQA — table math	0.019	0.082	0.126
ConvFinQA — multi-turn math	0.148	0.146	0.203
FPB — sentiment	0.812	0.790	0.794
Headlines — classification	0.745	0.714	0.721
NER — extraction (F1)	0.238	0.150	0.180
FiQA-SA — aspect sentiment	0.681	0.319	0.370
Average	0.441	0.367	0.399

Reading this: CPT alone improved FinQA numerical reasoning (0.019 → 0.082) but lowered the average — continued pre-training on raw text sharpened reasoning while disrupting classification calibration. The following SFT stage recovers part of the gap. See the final model card for the full pipeline, datasets, benchmarks, and analysis.

Training details

Base: Qwen/Qwen3-14B-Base (Apache-2.0)
Method: QLoRA CPT (r=64, α=128), packed seq len 2048, effective batch 128, lr 2e-4 cosine, 800 steps, Flash-Attention 2.
Hardware: 8× A100 SXM4 40 GB (AWS p4d.24xlarge), ~5 h.

Not financial advice. Verify all outputs.

Downloads last month: 7

Safetensors

Model size

15B params

Tensor type

BF16

Model tree for maxint-inc/Finsight-Qwen3-14B-CPT

Base model

Qwen/Qwen3-14B-Base

Finetuned

(92)

this model

Finsight — Qwen3-14B Finance (CPT only)

What this stage did

Results (on the FinBen flare_* finance benchmarks)

Training details

Model tree for maxint-inc/Finsight-Qwen3-14B-CPT

Results (on the FinBen `flare_*` finance benchmarks)