BGE v1.5

app.py

@@ -279,10 +279,10 @@ EXTERNAL_MODELS = [
     "allenai-specter",
     "bert-base-swedish-cased",
     "bert-base-uncased",
-    "bge-base-zh",
-    "bge-large-zh",
+    "bge-base-zh-v1.5",
+    "bge-large-zh-v1.5",
     "bge-large-zh-noinstruct",
-    "bge-small-zh",    
+    "bge-small-zh-v1.5",
     "contriever-base-msmarco",
     "cross-en-de-roberta-sentence-transformer",
     "dfm-encoder-large-v1",
@@ -355,10 +355,10 @@ EXTERNAL_MODEL_TO_LINK = {
     "all-mpnet-base-v2": "https://huggingface.co/sentence-transformers/all-mpnet-base-v2",
     "bert-base-swedish-cased": "https://huggingface.co/KB/bert-base-swedish-cased",
     "bert-base-uncased": "https://huggingface.co/bert-base-uncased",
-    "bge-base-zh": "https://huggingface.co/BAAI/bge-base-zh",
-    "bge-large-zh": "https://huggingface.co/BAAI/bge-large-zh",
+    "bge-base-zh-v1.5": "https://huggingface.co/BAAI/bge-base-zh-v1.5",
+    "bge-large-zh-v1.5": "https://huggingface.co/BAAI/bge-large-zh-v1.5",
     "bge-large-zh-noinstruct": "https://huggingface.co/BAAI/bge-large-zh-noinstruct",
-    "bge-small-zh": "https://huggingface.co/BAAI/bge-small-zh",
+    "bge-small-zh-v1.5": "https://huggingface.co/BAAI/bge-small-zh-v1.5",
     "contriever-base-msmarco": "https://huggingface.co/nthakur/contriever-base-msmarco",
     "cross-en-de-roberta-sentence-transformer": "https://huggingface.co/T-Systems-onsite/cross-en-de-roberta-sentence-transformer",
     "DanskBERT": "https://huggingface.co/vesteinn/DanskBERT",
@@ -431,10 +431,10 @@ EXTERNAL_MODEL_TO_DIM = {
     "allenai-specter": 768,    
     "bert-base-swedish-cased": 768,
     "bert-base-uncased": 768,
-    "bge-base-zh": 768,
-    "bge-large-zh": 1024,
+    "bge-base-zh-v1.5": 768,
+    "bge-large-zh-v1.5": 1024,
     "bge-large-zh-noinstruct": 1024,
-    "bge-small-zh": 512,
+    "bge-small-zh-v1.5": 512,
     "contriever-base-msmarco": 768,
     "cross-en-de-roberta-sentence-transformer": 768,
     "DanskBERT": 768,
@@ -507,10 +507,10 @@ EXTERNAL_MODEL_TO_SEQLEN = {
     "allenai-specter": 512,
     "bert-base-swedish-cased": 512,    
     "bert-base-uncased": 512,
-    "bge-base-zh": 512,
-    "bge-large-zh": 512,
+    "bge-base-zh-v1.5": 512,
+    "bge-large-zh-v1.5": 512,
     "bge-large-zh-noinstruct": 512,
-    "bge-small-zh": 512,    
+    "bge-small-zh-v1.5": 512,    
     "contriever-base-msmarco": 512,
     "cross-en-de-roberta-sentence-transformer": 514,
     "DanskBERT": 514,
@@ -583,10 +583,10 @@ EXTERNAL_MODEL_TO_SIZE = {
     "all-mpnet-base-v2": 0.44,
     "bert-base-uncased": 0.44,
     "bert-base-swedish-cased": 0.50,
-    "bge-base-zh": 0.41,
-    "bge-large-zh": 1.30,
+    "bge-base-zh-v1.5": 0.41,
+    "bge-large-zh-v1.5": 1.30,
     "bge-large-zh-noinstruct": 1.30,
-    "bge-small-zh": 0.10,      
+    "bge-small-zh-v1.5": 0.10,      
     "cross-en-de-roberta-sentence-transformer": 1.11,
     "contriever-base-msmarco": 0.44,
     "DanskBERT": 0.50,
@@ -675,6 +675,9 @@ MODELS_TO_SKIP = {
     "kozistr/fused-large-en",
     "sionic-ai/sionic-ai-v2", # Wait for https://huggingface.co/sionic-ai/sionic-ai-v2/discussions/1
     "sionic-ai/sionic-ai-v1", # Wait for https://huggingface.co/sionic-ai/sionic-ai-v2/discussions/1
+    "BAAI/bge-large-en", # Deprecated in favor of v1.5
+    "BAAI/bge-base-en", # Deprecated in favor of v1.5
+    "BAAI/bge-small-en", # Deprecated in favor of v1.5
 }
 
 EXTERNAL_MODEL_RESULTS = {model: {k: {v: []} for k, v in TASK_TO_METRIC.items()} for model in EXTERNAL_MODELS}