Training in progress, step 301

chat_template.jinja

@@ -1,10 +1 @@
-
-{%- for message in messages %}
-    {%- if message['role'] == 'system' %}
-        {{ '<|system|>' + ' ' + message['content'].strip() }}
-    {%- elif message['role'] == 'user' %}
-        {{ '<|instruction|>' + ' ' + message['content'].strip() }}
-    {%- elif message['role'] == 'assistant' %}
-        {{ '<|response|>' + ' ' + message['content'].strip() }}
-    {%- endif %}
-{%- endfor %}
+{{ bos_token }}{% for message in messages %}{% if message['role'] == 'user' %}{{ '<|start_header_id|>user<|end_header_id|> ' + message['content'] + ' <|eot_id|>' }}{% elif message['role'] == 'assistant' %}{{ '<|start_header_id|>assistant<|end_header_id|> ' + message['content'] + ' <|eot_id|>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|> ' }}{% endif %}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c9c6b17d8f5bd984411adb4bcb41a5a969e0516cfb4e11da8c4c9a4afec8bad4
+oid sha256:dff691ae8c26f29915c826edf637c3238a0409808304a965811c84b6d215c76e
 size 2013432032

special_tokens_map.json

@@ -6,14 +6,8 @@
     "rstrip": false,
     "single_word": false
   },
-  "eos_token": {
-    "content": "<|endoftext|>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "pad_token": "<|endoftext|>",
+  "eos_token": "<|eot_id|>",
+  "pad_token": "<|eot_id|>",
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,

tokenizer.json

@@ -2,7 +2,7 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 512,
+    "max_length": 256,
     "strategy": "LongestFirst",
     "stride": 0
   },

tokenizer_config.json

@@ -83,11 +83,15 @@
   },
   "bos_token": "<sos>",
   "clean_up_tokenization_spaces": true,
-  "eos_token": "<|endoftext|>",
+  "eos_token": "<|eot_id|>",
   "extra_special_tokens": {},
+  "max_length": 512,
   "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "<|endoftext|>",
+  "pad_token": "<|eot_id|>",
+  "stride": 0,
   "tokenizer_class": "PreTrainedTokenizer",
   "trim_offsets": true,
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
   "unk_token": "<unk>"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f6851789ee89c985cc7b366590e9859207c38763ef73f86b33c1ca7dd65b5b5b
+oid sha256:2916ab9e1755803b1193e934ee58981fcce2b041c1bb51d121df465f7d8fd9b2
 size 5240