初始化项目,由ModelHub XC社区提供模型
Model: NeshVerse/Flash-financial-analysis-lfm-1.2b Source: Original Platform
This commit is contained in:
2
.gitattributes
vendored
Normal file
2
.gitattributes
vendored
Normal file
@@ -0,0 +1,2 @@
|
|||||||
|
model.safetensors filter=lfs diff=lfs merge=lfs -text
|
||||||
|
flash-financial-analysis-q8_0-v2.gguf filter=lfs diff=lfs merge=lfs -text
|
||||||
109
README.md
Normal file
109
README.md
Normal file
@@ -0,0 +1,109 @@
|
|||||||
|
---
|
||||||
|
license: apache-2.0
|
||||||
|
language:
|
||||||
|
- en
|
||||||
|
base_model: LiquidAI/LFM2.5-1.2B-Base
|
||||||
|
tags:
|
||||||
|
- finance
|
||||||
|
- sales-analysis
|
||||||
|
- structured-data
|
||||||
|
- unsloth
|
||||||
|
- lora
|
||||||
|
- lfm
|
||||||
|
- data-analysis
|
||||||
|
- time-series
|
||||||
|
- business-intelligence
|
||||||
|
- fast-inference
|
||||||
|
- efficient-finetuning
|
||||||
|
pipeline_tag: text-generation
|
||||||
|
library_name: transformers
|
||||||
|
---
|
||||||
|
|
||||||
|
# ⚡ Flash-Financial-Analysis-LFM-1.2B
|
||||||
|
|
||||||
|
**Lightning-fast financial intelligence for structured data analysis**
|
||||||
|
|
||||||
|
A blazing-fast, customized, lightweight language model optimized for real-time sales & stock analytics, inventory insights, and financial reporting based on the LiquidAI 1.2B base model supervised fine-tuned FP16 model.
|
||||||
|
|
||||||
|
## Rag Model Github Repo:
|
||||||
|
https://github.com/neshverse/Flash-RAG-web-GUI/tree/main
|
||||||
|
|
||||||
|
## Space (Real-time Testing)
|
||||||
|
https://huggingface.co/spaces/NeshVerse/Flash-financial-analysis-lfm-1.2b
|
||||||
|
|
||||||
|
Note: If you get any error while chatting, just refresh the page.
|
||||||
|
|
||||||
|
## Model Details
|
||||||
|
|
||||||
|
| Attribute | Value |
|
||||||
|
|-----------|-------|
|
||||||
|
| **Base Architecture** | [LiquidAI/LFM2.5-1.2B-Base](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Base) |
|
||||||
|
| **Fine-tuning** | LoRA (r=4, alpha=8) |
|
||||||
|
| **Context Window** | 1,024 tokens |
|
||||||
|
| **Precision** | FP16 |
|
||||||
|
| **Parameters** | 1.2B base + ~500K LoRA |
|
||||||
|
|
||||||
|
## Training Summary
|
||||||
|
|
||||||
|
- **Total Samples**: 39,435 (37,463 train / 1,972 validation)
|
||||||
|
- **Training Duration**: 2.4 hours
|
||||||
|
- **Final Loss**: 0.497 (train) / 0.508 (validation)
|
||||||
|
- **Hardware**: Consumer GPU (T4)
|
||||||
|
|
||||||
|
## Capabilities
|
||||||
|
|
||||||
|
- **Sales Analytics**: Real-time sales data querying and analysis
|
||||||
|
- **Stock Analytics**: Inventory levels, turnover rates, stock movement
|
||||||
|
- **Financial Reporting**: Automated report generation from structured data
|
||||||
|
- **Inventory Insights**: Product performance, seasonal trends, demand forecasting
|
||||||
|
|
||||||
|
## Performance
|
||||||
|
|
||||||
|
- **Inference Speed**: ~0.55 it/s (T4 GPU)
|
||||||
|
- **Memory Usage**: ~6GB (4-bit loaded)
|
||||||
|
- **Batch Size**: 4 (effective 8 with grad accum)
|
||||||
|
- **Max Sequence**: 1,024 tokens
|
||||||
|
|
||||||
|
## Limitations
|
||||||
|
|
||||||
|
- Optimized for structured financial/sales data queries
|
||||||
|
- Context window limited to 1,024 tokens
|
||||||
|
- Training data from 2022-2023; may not reflect current market conditions
|
||||||
|
- Best performance on English language inputs
|
||||||
|
|
||||||
|
## Model Files
|
||||||
|
|
||||||
|
| File | Format | Size | Description | Use Case |
|
||||||
|
|------|--------|------|-------------|----------|
|
||||||
|
| `model.safetensors` | FP16 | ~2.4 GB | Original full precision | Maximum quality, GPU inference |
|
||||||
|
| `flash-financial-analysis-q8_0.gguf` | Q8_0 | ~1.2 GB | 8-bit quantized (llama.cpp) | CPU inference, Ollama, LM Studio |
|
||||||
|
|
||||||
|
## Quantized Version (Q8_0)
|
||||||
|
|
||||||
|
We now provide a **Q8_0 quantized version** for easier deployment:
|
||||||
|
|
||||||
|
- **Format**: GGUF (llama.cpp compatible)
|
||||||
|
- **Size**: ~50% smaller than FP16 (1.2 GB vs 2.4 GB)
|
||||||
|
- **Quality**: ~99.9% of original performance
|
||||||
|
- **Tools**: Works with llama.cpp, Ollama, LM Studio, llama-cpp-python
|
||||||
|
|
||||||
|
### Download Q8_0
|
||||||
|
|
||||||
|
```bash
|
||||||
|
# Using huggingface-cli
|
||||||
|
huggingface-cli download NeshVerse/Flash-financial-analysis-lfm-1.2b flash-financial-analysis-q8_0.gguf
|
||||||
|
|
||||||
|
# Or direct download
|
||||||
|
wget https://huggingface.co/NeshVerse/Flash-financial-analysis-lfm-1.2b/resolve/main/flash-financial-analysis-q8_0.gguf
|
||||||
|
|
||||||
|
## Quick Start
|
||||||
|
|
||||||
|
```python
|
||||||
|
from unsloth import FastLanguageModel
|
||||||
|
|
||||||
|
model, tokenizer = FastLanguageModel.from_pretrained(
|
||||||
|
"NeshVerse/Flash-financial-analysis-lfm-1.2b",
|
||||||
|
max_seq_length=1024,
|
||||||
|
load_in_4bit=True,
|
||||||
|
trust_remote_code=True,
|
||||||
|
)
|
||||||
7
chat_template.jinja
Normal file
7
chat_template.jinja
Normal file
@@ -0,0 +1,7 @@
|
|||||||
|
{{- bos_token -}}{%- set system_prompt = "" -%}{%- set ns = namespace(system_prompt="") -%}{%- if messages[0]["role"] == "system" -%} {%- set ns.system_prompt = messages[0]["content"] -%} {%- set messages = messages[1:] -%}{%- endif -%}{%- if tools -%} {%- set ns.system_prompt = ns.system_prompt + ("
|
||||||
|
" if ns.system_prompt else "") + "List of tools: <|tool_list_start|>[" -%} {%- for tool in tools -%} {%- if tool is not string -%} {%- set tool = tool | tojson -%} {%- endif -%} {%- set ns.system_prompt = ns.system_prompt + tool -%} {%- if not loop.last -%} {%- set ns.system_prompt = ns.system_prompt + ", " -%} {%- endif -%} {%- endfor -%} {%- set ns.system_prompt = ns.system_prompt + "]<|tool_list_end|>" -%}{%- endif -%}{%- if ns.system_prompt -%} {{- "<|im_start|>system
|
||||||
|
" + ns.system_prompt + "<|im_end|>
|
||||||
|
" -}}{%- endif -%}{%- for message in messages -%} {{- "<|im_start|>" + message["role"] + "
|
||||||
|
" -}} {%- set content = message["content"] -%} {%- if content is not string -%} {%- set content = content | tojson -%} {%- endif -%} {%- if message["role"] == "tool" -%} {%- set content = "<|tool_response_start|>" + content + "<|tool_response_end|>" -%} {%- endif -%} {{- content + "<|im_end|>
|
||||||
|
" -}}{%- endfor -%}{%- if add_generation_prompt -%} {{- "<|im_start|>assistant
|
||||||
|
" -}}{%- endif -%}
|
||||||
62
config.json
Normal file
62
config.json
Normal file
@@ -0,0 +1,62 @@
|
|||||||
|
{
|
||||||
|
"architectures": [
|
||||||
|
"Lfm2ForCausalLM"
|
||||||
|
],
|
||||||
|
"block_auto_adjust_ff_dim": true,
|
||||||
|
"block_dim": 2048,
|
||||||
|
"block_ff_dim": 12288,
|
||||||
|
"block_ffn_dim_multiplier": 1.0,
|
||||||
|
"block_mlp_init_scale": 1.0,
|
||||||
|
"block_multiple_of": 256,
|
||||||
|
"block_norm_eps": 1e-05,
|
||||||
|
"block_out_init_scale": 1.0,
|
||||||
|
"block_use_swiglu": true,
|
||||||
|
"block_use_xavier_init": true,
|
||||||
|
"bos_token_id": 1,
|
||||||
|
"conv_L_cache": 3,
|
||||||
|
"conv_bias": false,
|
||||||
|
"conv_dim": 2048,
|
||||||
|
"conv_use_xavier_init": true,
|
||||||
|
"dtype": "float16",
|
||||||
|
"eos_token_id": 7,
|
||||||
|
"hidden_size": 2048,
|
||||||
|
"initializer_range": 0.02,
|
||||||
|
"intermediate_size": 12288,
|
||||||
|
"layer_types": [
|
||||||
|
"conv",
|
||||||
|
"conv",
|
||||||
|
"full_attention",
|
||||||
|
"conv",
|
||||||
|
"conv",
|
||||||
|
"full_attention",
|
||||||
|
"conv",
|
||||||
|
"conv",
|
||||||
|
"full_attention",
|
||||||
|
"conv",
|
||||||
|
"full_attention",
|
||||||
|
"conv",
|
||||||
|
"full_attention",
|
||||||
|
"conv",
|
||||||
|
"full_attention",
|
||||||
|
"conv"
|
||||||
|
],
|
||||||
|
"max_position_embeddings": 128000,
|
||||||
|
"model_type": "lfm2",
|
||||||
|
"norm_eps": 1e-05,
|
||||||
|
"num_attention_heads": 32,
|
||||||
|
"num_heads": 32,
|
||||||
|
"num_hidden_layers": 16,
|
||||||
|
"num_key_value_heads": 8,
|
||||||
|
"pad_token_id": 0,
|
||||||
|
"rope_parameters": {
|
||||||
|
"rope_theta": 1000000.0,
|
||||||
|
"rope_type": "default"
|
||||||
|
},
|
||||||
|
"tie_embedding": true,
|
||||||
|
"tie_word_embeddings": true,
|
||||||
|
"transformers_version": "5.1.0",
|
||||||
|
"unsloth_version": "2026.2.1",
|
||||||
|
"use_cache": true,
|
||||||
|
"use_pos_enc": true,
|
||||||
|
"vocab_size": 65536
|
||||||
|
}
|
||||||
3
flash-financial-analysis-q8_0-v2.gguf
Normal file
3
flash-financial-analysis-q8_0-v2.gguf
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:d58d2595df10e711baeb916ce39bedb90d894aa6f5e3592f376699f8936d924d
|
||||||
|
size 1246252928
|
||||||
8
generation_config.json
Normal file
8
generation_config.json
Normal file
@@ -0,0 +1,8 @@
|
|||||||
|
{
|
||||||
|
"_from_model_config": true,
|
||||||
|
"bos_token_id": 1,
|
||||||
|
"eos_token_id": 7,
|
||||||
|
"max_length": 128000,
|
||||||
|
"pad_token_id": 0,
|
||||||
|
"transformers_version": "5.1.0"
|
||||||
|
}
|
||||||
3
model.safetensors
Normal file
3
model.safetensors
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:36bc54c83905c9157aa3739ba6966954d3487f6e554ebfc993d0674827638096
|
||||||
|
size 2340697784
|
||||||
323812
tokenizer.json
Normal file
323812
tokenizer.json
Normal file
File diff suppressed because it is too large
Load Diff
20
tokenizer_config.json
Normal file
20
tokenizer_config.json
Normal file
@@ -0,0 +1,20 @@
|
|||||||
|
{
|
||||||
|
"backend": "tokenizers",
|
||||||
|
"bos_token": "<|startoftext|>",
|
||||||
|
"clean_up_tokenization_spaces": false,
|
||||||
|
"eos_token": "<|im_end|>",
|
||||||
|
"is_local": false,
|
||||||
|
"legacy": false,
|
||||||
|
"model_input_names": [
|
||||||
|
"input_ids",
|
||||||
|
"attention_mask"
|
||||||
|
],
|
||||||
|
"model_max_length": 1000000000000000019884624838656,
|
||||||
|
"pad_token": "<|pad|>",
|
||||||
|
"padding_side": "left",
|
||||||
|
"sp_model_kwargs": {},
|
||||||
|
"spaces_between_special_tokens": false,
|
||||||
|
"tokenizer_class": "TokenizersBackend",
|
||||||
|
"use_default_system_prompt": false,
|
||||||
|
"use_fast": true
|
||||||
|
}
|
||||||
Reference in New Issue
Block a user