Update pattern_analyzer.py

pattern_analyzer.py

@@ -1,5 +1,6 @@
 import os
 os.environ['HF_HOME'] = '/tmp/huggingface'
+
 from transformers import AutoModelForCausalLM, AutoTokenizer
 import torch
 import json
@@ -12,20 +13,28 @@ class PatternAnalyzer:
             "device_map": "auto",
             "torch_dtype": torch.float32,
             "low_cpu_mem_usage": True,
-            "max_memory": {"cpu": "8GB"},
-            "offload_folder": "/tmp/offload"
+            "max_memory": {
+                "cpu": "4GB",
+                "disk": "8GB"
+            },
+            "offload_folder": "/tmp/offload",
+            "load_in_8bit": True,
+            "revision": "main"
         }
-        
+
         self.model = AutoModelForCausalLM.from_pretrained(
             "tmmdev/codellama-pattern-analysis",
-            **model_kwargs
+            **model_kwargs,
+            use_safetensors=True,
+            trust_remote_code=True,
+            resume_download=True
         )
-        
+
         self.tokenizer = AutoTokenizer.from_pretrained(
             "tmmdev/codellama-pattern-analysis",
             use_fast=True
         )
-        
+
         self.basic_patterns = {
             'channel': {'min_points': 4, 'confidence_threshold': 0.7},
             'triangle': {'min_points': 3, 'confidence_threshold': 0.75},
@@ -34,6 +43,7 @@ class PatternAnalyzer:
             'double_top': {'max_deviation': 0.02, 'confidence_threshold': 0.85},
             'double_bottom': {'max_deviation': 0.02, 'confidence_threshold': 0.85}
         }
+
         self.pattern_logic = PatternLogic()
 
     def analyze_data(self, ohlcv_data):
@@ -44,16 +54,13 @@ class PatternAnalyzer:
         2. Triangle: Must have clear convergence point
         3. Support: Minimum 3 price bounces
         4. Resistance: Minimum 3 price rejections
-
         INPUT DATA:
         {ohlcv_data.to_json(orient='records')}
-
         Return ONLY high-confidence patterns (>0.8) in JSON format with exact price coordinates."""
 
         inputs = self.tokenizer(data_prompt, return_tensors="pt")
         outputs = self.model.generate(**inputs, max_length=1000)
         analysis = self.tokenizer.decode(outputs[0])
-
         return self.parse_analysis(analysis)
 
     def parse_analysis(self, analysis_text):
@@ -61,10 +68,9 @@ class PatternAnalyzer:
             json_start = analysis_text.find('{')
             json_end = analysis_text.rfind('}') + 1
             json_str = analysis_text[json_start:json_end]
-            
             analysis_data = json.loads(json_str)
             patterns = []
-            
+
             for pattern in analysis_data.get('patterns', []):
                 pattern_type = pattern.get('type')
                 if pattern_type in self.basic_patterns:
@@ -79,8 +85,6 @@ class PatternAnalyzer:
                                 'timestamp': pd.Timestamp.now().isoformat()
                             }
                         })
-            
             return patterns
-            
         except json.JSONDecodeError:
             return []