More models

evals/models.py

@@ -15,12 +15,19 @@ models = [
     "openai/gpt-4o-mini",  # 0.6$/M tokens
     # "anthropic/claude-3.5-haiku", # 4$/M tokens -> too expensive for dev
     "meta-llama/llama-3.3-70b-instruct",  # 0.3$/M tokens
+    "meta-llama/llama-3.1-70b-instruct",  # 0.3$/M tokens
+    "meta-llama/llama-3-70b-instruct", # 0.4$/M tokens
     "mistralai/mistral-small-24b-instruct-2501",  # 0.14$/M tokens
+    "mistralai/mistral-nemo",
     "google/gemini-2.0-flash-001",  # 0.4$/M tokens
+    "google/gemini-2.0-flash-lite-001",  # 0.3$/M tokens
+    "google/gemma-3-27b-it",  # 0.2$/M tokens
     # "qwen/qwen-turbo", # 0.2$/M tokens; recognizes "inappropriate content"
-    # "deepseek/deepseek-chat",  # 0.9$/M tokens
+    "qwen/qwq-32b",
+    # "deepseek/deepseek-chat",  # 1.3$/M tokens
     # "microsoft/phi-4",  # 0.07$/M tokens; only 16k tokens context
-    "google/gemma-3-27b-it",  # 0.2$/M tokens
+    "microsoft/phi-4-multimodal-instruct",
+    "amazon/nova-micro-v1",  # 0.09$/M tokens
 ]
 model_fast = "meta-llama/llama-3.3-70b-instruct"
 

frontend/public/results.json

@@ -18,6 +18,21 @@
     {
       "rank": 2,
       "provider": "Google",
+      "model": "Gemini 2.0 Flash Lite 001",
+      "hf_id": null,
+      "creation_date": null,
+      "size": null,
+      "type": "Commercial",
+      "license": null,
+      "average": 0.66,
+      "classification_accuracy": 0.73,
+      "language_modeling_chrf": 0.97,
+      "translation_bleu": 0.4,
+      "translation_chrf": 0.54
+    },
+    {
+      "rank": 3,
+      "provider": "Google",
       "model": "Gemma 3 27b It",
       "hf_id": "google/gemma-3-27b-it",
       "creation_date": "2025-03-01",
@@ -31,7 +46,52 @@
       "translation_chrf": 0.54
     },
     {
-      "rank": 3,
+      "rank": 4,
+      "provider": "Meta Llama",
+      "model": "Llama 3.1 70b Instruct",
+      "hf_id": "meta-llama/Llama-3.1-70B-Instruct",
+      "creation_date": "2024-07-16",
+      "size": 70553706496.0,
+      "type": "Open",
+      "license": "Llama3.1",
+      "average": 0.62,
+      "classification_accuracy": 0.57,
+      "language_modeling_chrf": 0.92,
+      "translation_bleu": 0.43,
+      "translation_chrf": 0.57
+    },
+    {
+      "rank": 5,
+      "provider": "Amazon",
+      "model": "Nova Micro V1",
+      "hf_id": null,
+      "creation_date": null,
+      "size": null,
+      "type": "Commercial",
+      "license": null,
+      "average": 0.61,
+      "classification_accuracy": 0.52,
+      "language_modeling_chrf": 0.94,
+      "translation_bleu": 0.4,
+      "translation_chrf": 0.56
+    },
+    {
+      "rank": 6,
+      "provider": "Meta Llama",
+      "model": "Llama 3 70b Instruct",
+      "hf_id": null,
+      "creation_date": null,
+      "size": null,
+      "type": "Commercial",
+      "license": null,
+      "average": 0.61,
+      "classification_accuracy": 0.8,
+      "language_modeling_chrf": 0.95,
+      "translation_bleu": 0.25,
+      "translation_chrf": 0.43
+    },
+    {
+      "rank": 7,
       "provider": "OpenAI",
       "model": "GPT 4o Mini",
       "hf_id": null,
@@ -46,7 +106,7 @@
       "translation_chrf": 0.55
     },
     {
-      "rank": 4,
+      "rank": 8,
       "provider": "MistralAI",
       "model": "Mistral Small 24b Instruct 2501",
       "hf_id": "mistralai/Mistral-Small-24B-Instruct-2501",
@@ -61,7 +121,7 @@
       "translation_chrf": 0.52
     },
     {
-      "rank": 5,
+      "rank": 9,
       "provider": "Meta Llama",
       "model": "Llama 3.3 70b Instruct",
       "hf_id": "meta-llama/Llama-3.3-70B-Instruct",
@@ -74,6 +134,51 @@
       "language_modeling_chrf": 0.94,
       "translation_bleu": 0.31,
       "translation_chrf": 0.48
+    },
+    {
+      "rank": 10,
+      "provider": "MistralAI",
+      "model": "Mistral Nemo",
+      "hf_id": null,
+      "creation_date": null,
+      "size": null,
+      "type": "Commercial",
+      "license": null,
+      "average": 0.55,
+      "classification_accuracy": 0.5,
+      "language_modeling_chrf": 0.88,
+      "translation_bleu": 0.32,
+      "translation_chrf": 0.49
+    },
+    {
+      "rank": 11,
+      "provider": "Microsoft",
+      "model": "Phi 4 Multimodal Instruct",
+      "hf_id": "microsoft/Phi-4-multimodal-instruct",
+      "creation_date": "2025-02-24",
+      "size": 5574460384.0,
+      "type": "Open",
+      "license": "Mit",
+      "average": 0.52,
+      "classification_accuracy": 0.42,
+      "language_modeling_chrf": 0.87,
+      "translation_bleu": 0.32,
+      "translation_chrf": 0.46
+    },
+    {
+      "rank": 12,
+      "provider": "Qwen",
+      "model": "Qwq 32b",
+      "hf_id": "Qwen/QwQ-32B",
+      "creation_date": "2025-03-05",
+      "size": 32763876352.0,
+      "type": "Open",
+      "license": "Apache 2.0",
+      "average": 0.25,
+      "classification_accuracy": 0.0,
+      "language_modeling_chrf": 0.48,
+      "translation_bleu": 0.21,
+      "translation_chrf": 0.3
     }
   ],
   "language_table": [
@@ -81,25 +186,13 @@
       "language_name": "English",
       "speakers": 1636485840,
       "family": "Indo-European",
-      "average": 0.51,
+      "average": 0.47,
       "in_benchmark": true,
       "NaN": 0.0,
-      "classification_accuracy": 0.65,
-      "language_modeling_chrf": 0.94,
-      "translation_bleu": 0.43,
-      "translation_chrf": 0.55
-    },
-    {
-      "language_name": "Chinese",
-      "speakers": 1304678914,
-      "family": "Sino-Tibetan",
-      "average": 0.5,
-      "in_benchmark": true,
-      "NaN": 0.0,
-      "classification_accuracy": 0.65,
-      "language_modeling_chrf": 0.93,
-      "translation_bleu": 0.38,
-      "translation_chrf": 0.55
+      "classification_accuracy": 0.58,
+      "language_modeling_chrf": 0.92,
+      "translation_bleu": 0.37,
+      "translation_chrf": 0.49
     },
     {
       "language_name": "French",
@@ -113,6 +206,18 @@
       "translation_bleu": 0.32,
       "translation_chrf": 0.49
     },
+    {
+      "language_name": "Chinese",
+      "speakers": 1304678914,
+      "family": "Sino-Tibetan",
+      "average": 0.46,
+      "in_benchmark": true,
+      "NaN": 0.0,
+      "classification_accuracy": 0.55,
+      "language_modeling_chrf": 0.86,
+      "translation_bleu": 0.35,
+      "translation_chrf": 0.53
+    },
     {
       "language_name": "Hindi",
       "speakers": 546882144,

results.json

@@ -3,33 +3,61 @@
     {
       "task": "classification",
       "metric": "accuracy",
-      "score": 0.5722222222222223,
+      "score": 0.5427083333333333,
       "bcp_47": 10,
-      "model": 5
+      "model": 12
     },
     {
       "task": "language_modeling",
       "metric": "chrf",
-      "score": 0.9360730990265229,
+      "score": 0.9024222998985517,
       "bcp_47": 10,
-      "model": 5
+      "model": 12
     },
     {
       "task": "translation",
       "metric": "bleu",
-      "score": 0.3508793079301233,
+      "score": 0.34325329881872996,
       "bcp_47": 10,
-      "model": 5
+      "model": 12
     },
     {
       "task": "translation",
       "metric": "chrf",
-      "score": 0.5117214627054559,
+      "score": 0.49764810942023735,
       "bcp_47": 10,
-      "model": 5
+      "model": 12
     }
   ],
   "models": [
+    {
+      "model": "amazon/nova-micro-v1",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.5166666666666666,
+      "bcp_47": 2
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9446198732700857,
+      "bcp_47": 2
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.40042093531509637,
+      "bcp_47": 2
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5642142196700637,
+      "bcp_47": 2
+    },
     {
       "model": "google/gemini-2.0-flash-001",
       "task": "classification",
@@ -58,6 +86,34 @@
       "score": 0.5828490054615683,
       "bcp_47": 2
     },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.7333333333333333,
+      "bcp_47": 2
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9710194350890375,
+      "bcp_47": 2
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.40085159165111883,
+      "bcp_47": 2
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5422821788946908,
+      "bcp_47": 2
+    },
     {
       "model": "google/gemma-3-27b-it",
       "task": "classification",
@@ -86,6 +142,62 @@
       "score": 0.5376336154503363,
       "bcp_47": 2
     },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.8,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9452435586756014,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.25148401884229143,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.4285750600098188,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.5666666666666667,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9203465184571391,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.43182300663190504,
+      "bcp_47": 2
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5679592059634284,
+      "bcp_47": 2
+    },
     {
       "model": "meta-llama/llama-3.3-70b-instruct",
       "task": "classification",
@@ -114,6 +226,62 @@
       "score": 0.4836914110309717,
       "bcp_47": 10
     },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.4166666666666667,
+      "bcp_47": 2
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.8700000415175042,
+      "bcp_47": 2
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.31733056990581465,
+      "bcp_47": 2
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.45631576469060464,
+      "bcp_47": 2
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.5,
+      "bcp_47": 2
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.8815544644693022,
+      "bcp_47": 2
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.3177444138044378,
+      "bcp_47": 2
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.49319228717306784,
+      "bcp_47": 2
+    },
     {
       "model": "mistralai/mistral-small-24b-instruct-2501",
       "task": "classification",
@@ -169,6 +337,34 @@
       "metric": "chrf",
       "score": 0.5452510379336759,
       "bcp_47": 2
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.0,
+      "bcp_47": 2
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.4813150156594517,
+      "bcp_47": 2
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.2144844735779058,
+      "bcp_47": 2
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.30433786997302065,
+      "bcp_47": 2
     }
   ],
   "languages": [
@@ -2554,8 +2750,8 @@
       "in_benchmark": true,
       "task": "classification",
       "metric": "accuracy",
-      "score": 0.6466666666666667,
-      "model": 5.0
+      "score": 0.5777777777777778,
+      "model": 12.0
     },
     {
       "bcp_47": "en",
@@ -2569,8 +2765,8 @@
       "in_benchmark": true,
       "task": "language_modeling",
       "metric": "chrf",
-      "score": 0.9391757502550891,
-      "model": 5.0
+      "score": 0.9222343234934963,
+      "model": 12.0
     },
     {
       "bcp_47": "en",
@@ -2584,8 +2780,8 @@
       "in_benchmark": true,
       "task": "translation",
       "metric": "bleu",
-      "score": 0.4274883186793429,
-      "model": 5.0
+      "score": 0.37035746903842287,
+      "model": 12.0
     },
     {
       "bcp_47": "en",
@@ -2599,8 +2795,8 @@
       "in_benchmark": true,
       "task": "translation",
       "metric": "chrf",
-      "score": 0.5456627250056182,
-      "model": 5.0
+      "score": 0.4880916692700535,
+      "model": 12.0
     },
     {
       "bcp_47": "eo",
@@ -10699,8 +10895,8 @@
       "in_benchmark": true,
       "task": "classification",
       "metric": "accuracy",
-      "score": 0.6466666666666667,
-      "model": 5.0
+      "score": 0.5499999999999999,
+      "model": 12.0
     },
     {
       "bcp_47": "zh",
@@ -10714,8 +10910,8 @@
       "in_benchmark": true,
       "task": "language_modeling",
       "metric": "chrf",
-      "score": 0.93253470927813,
-      "model": 5.0
+      "score": 0.8599948525016986,
+      "model": 12.0
     },
     {
       "bcp_47": "zh",
@@ -10729,8 +10925,8 @@
       "in_benchmark": true,
       "task": "translation",
       "metric": "bleu",
-      "score": 0.37546421356438325,
-      "model": 5.0
+      "score": 0.3532292543512247,
+      "model": 12.0
     },
     {
       "bcp_47": "zh",
@@ -10744,8 +10940,8 @@
       "in_benchmark": true,
       "task": "translation",
       "metric": "chrf",
-      "score": 0.553563744610482,
-      "model": 5.0
+      "score": 0.529398790799104,
+      "model": 12.0
     },
     {
       "bcp_47": "zmi",
@@ -10794,6 +10990,70 @@
     }
   ],
   "scores": [
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.5333333333333333,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9725001956658679,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.4491277841667736,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5740458676508566,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.5,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9167395508743035,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.3517140864634192,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "amazon/nova-micro-v1",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5543825716892707,
+      "sentence_nr": 14.5
+    },
     {
       "model": "google/gemini-2.0-flash-001",
       "bcp_47": "en",
@@ -10858,6 +11118,70 @@
       "score": 0.5606266861920302,
       "sentence_nr": 14.5
     },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.7333333333333333,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.990925430282282,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.37911136698810943,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5094402087357145,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.7333333333333333,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9511134398957932,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.4225918163141283,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "google/gemini-2.0-flash-lite-001",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5751241490536672,
+      "sentence_nr": 14.5
+    },
     {
       "model": "google/gemma-3-27b-it",
       "bcp_47": "en",
@@ -10922,6 +11246,134 @@
       "score": 0.520771580386218,
       "sentence_nr": 14.5
     },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.8333333333333334,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9674315682816375,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.18722412351358647,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.34151371128305424,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.7666666666666667,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9230555490695652,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.3157439141709964,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3-70b-instruct",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5156364087365835,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.7,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9701295103188484,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.44443705644214526,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5485685299214524,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.43333333333333335,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.8705635265954298,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.4192089568216648,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "meta-llama/llama-3.1-70b-instruct",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5873498820054043,
+      "sentence_nr": 14.5
+    },
     {
       "model": "meta-llama/llama-3.3-70b-instruct",
       "bcp_47": "ar",
@@ -11242,6 +11694,134 @@
       "score": 0.5862284100611604,
       "sentence_nr": 14.5
     },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.43333333333333335,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9268050965065061,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.34049537977839345,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.4566714452688056,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.4,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.8131949865285024,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.2941657600332359,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "microsoft/phi-4-multimodal-instruct",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.4559600841124037,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.4666666666666667,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.9383955895073849,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.3057719571177098,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.45969934521843914,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.5333333333333333,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.8247133394312195,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.32971687049116577,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "mistralai/mistral-nemo",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.5266852291276966,
+      "sentence_nr": 14.5
+    },
     {
       "model": "mistralai/mistral-small-24b-instruct-2501",
       "bcp_47": "en",
@@ -11369,6 +11949,70 @@
       "metric": "chrf",
       "score": 0.559410465345808,
       "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "en",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.0,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "en",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.6047457400839834,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.20068036705764214,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "en",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.23884729813422853,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "zh",
+      "task": "classification",
+      "metric": "accuracy",
+      "score": 0.0,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "zh",
+      "task": "language_modeling",
+      "metric": "chrf",
+      "score": 0.35788429123492,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "bleu",
+      "score": 0.22828858009816946,
+      "sentence_nr": 14.5
+    },
+    {
+      "model": "qwen/qwq-32b",
+      "bcp_47": "zh",
+      "task": "translation",
+      "metric": "chrf",
+      "score": 0.3698284418118128,
+      "sentence_nr": 14.5
     }
   ]
 }