add Mistral evals

evals/arc/arc_nl_Mistral-7B-v0.1.json

@@ -1,10 +1,10 @@
 {
   "results": {
     "arc_nl": {
-      "acc": 0.41916167664670656,
-      "acc_stderr": 0.01443766613132885,
-      "acc_norm": 0.42172797262617623,
-      "acc_norm_stderr": 0.014449768025902365
+      "acc": 0.41146278870829767,
+      "acc_stderr": 0.014398950037131928,
+      "acc_norm": 0.42771599657827203,
+      "acc_norm_stderr": 0.014476451794175197
     }
   },
   "versions": {
@@ -12,8 +12,8 @@
   },
   "config": {
     "model": "hf-auto",
-    "model_args": "pretrained=mistralai/Mistral-7B-v0.1,use_accelerate=True,device_map_option=auto,dtype=bfloat16,load_in_8bit=True",
-    "batch_size": "auto",
+    "model_args": "pretrained=mistralai/Mistral-7B-v0.1,use_accelerate=True,device_map_option=auto,dtype=auto,load_in_8bit=True",
+    "batch_size": 2,
     "device": "cuda",
     "no_cache": false,
     "limit": null,

evals/hellaswag/hellaswag_nl_Mistral-7B-v0.1.json

@@ -0,0 +1,23 @@
+{
+  "results": {
+    "hellaswag_nl": {
+      "acc": 0.4418780356179169,
+      "acc_stderr": 0.005159605430771658,
+      "acc_norm": 0.5829465731246627,
+      "acc_norm_stderr": 0.005122841872653067
+    }
+  },
+  "versions": {
+    "hellaswag_nl": 1
+  },
+  "config": {
+    "model": "hf-auto",
+    "model_args": "pretrained=mistralai/Mistral-7B-v0.1,use_accelerate=True,device_map_option=auto,dtype=auto,load_in_8bit=True",
+    "batch_size": 2,
+    "device": "cuda",
+    "no_cache": false,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "description_dict": {}
+  }
+}

evals/truthfulqa/truthfulqa_nl_Mistral-7B-v0.1.json

@@ -12,7 +12,7 @@
   },
   "config": {
     "model": "hf-auto",
-    "model_args": "pretrained=mistralai/Mistral-7B-v0.1,use_accelerate=True,device_map_option=auto,dtype=bfloat16,load_in_8bit=True",
+    "model_args": "pretrained=mistralai/Mistral-7B-v0.1,use_accelerate=True,device_map_option=auto,dtype=auto,load_in_8bit=True",
     "batch_size": 2,
     "device": "cuda",
     "no_cache": false,