Upload Maaza-MLM-135M-JSON-v1 - v1.0.0 production release
Browse files
README.md
CHANGED
|
@@ -1,3 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# CycleCore Maaza MLM-135M-JSON v1.0.0
|
| 2 |
|
| 3 |
Micro Language Model (135M parameters) specialized for JSON extraction on edge devices.
|
|
@@ -227,8 +244,6 @@ print(result)
|
|
| 227 |
|
| 228 |
## Model Comparison
|
| 229 |
|
| 230 |
-
For guidance on choosing between MLM-135M and SLM-360M, see our [Model Comparison Guide](https://github.com/CycleCore/SLMBench/blob/main/docs/MODEL_COMPARISON.md).
|
| 231 |
-
|
| 232 |
**Quick Decision**:
|
| 233 |
- **Use MLM-135M** if: Ultra-low latency required, simple schemas (2-4 fields), <500MB deployment size
|
| 234 |
- **Use SLM-360M** if: Higher accuracy needed, medium/complex schemas, willing to use ~1GB deployment size
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
base_model: HuggingFaceTB/SmolLM2-135M
|
| 6 |
+
tags:
|
| 7 |
+
- json
|
| 8 |
+
- structured-output
|
| 9 |
+
- edge-ai
|
| 10 |
+
- iot
|
| 11 |
+
- micro-language-model
|
| 12 |
+
- peft
|
| 13 |
+
- lora
|
| 14 |
+
library_name: transformers
|
| 15 |
+
pipeline_tag: text-generation
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
# CycleCore Maaza MLM-135M-JSON v1.0.0
|
| 19 |
|
| 20 |
Micro Language Model (135M parameters) specialized for JSON extraction on edge devices.
|
|
|
|
| 244 |
|
| 245 |
## Model Comparison
|
| 246 |
|
|
|
|
|
|
|
| 247 |
**Quick Decision**:
|
| 248 |
- **Use MLM-135M** if: Ultra-low latency required, simple schemas (2-4 fields), <500MB deployment size
|
| 249 |
- **Use SLM-360M** if: Higher accuracy needed, medium/complex schemas, willing to use ~1GB deployment size
|