First commit: Upload model QG Simplified

config.json

@@ -1,24 +1,27 @@
 {
   "architectures": [
-    "T5ForConditionalGeneration"
+    "LongT5ForConditionalGeneration"
   ],
-  "classifier_dropout": 0.0,
-  "d_ff": 3072,
+  "bos_token_id": null,
+  "d_ff": 2048,
   "d_kv": 64,
   "d_model": 768,
   "decoder_start_token_id": 0,
-  "dense_act_fn": "relu",
+  "dense_act_fn": "gelu_new",
   "dropout_rate": 0.1,
   "dtype": "float32",
+  "encoder_attention_type": "transient-global",
   "eos_token_id": 1,
-  "feed_forward_proj": "relu",
+  "feed_forward_proj": "gated-gelu",
+  "global_block_size": 16,
   "initializer_factor": 1.0,
   "is_decoder": false,
   "is_encoder_decoder": true,
-  "is_gated_act": false,
+  "is_gated_act": true,
   "layer_norm_epsilon": 1e-06,
-  "model_type": "t5",
-  "n_positions": 512,
+  "local_radius": 127,
+  "model_type": "longt5",
+  "n_positions": 4096,
   "num_decoder_layers": 12,
   "num_heads": 12,
   "num_layers": 12,
@@ -26,37 +29,7 @@
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,
-  "scale_decoder_outputs": true,
-  "task_specific_params": {
-    "summarization": {
-      "early_stopping": true,
-      "length_penalty": 2.0,
-      "max_length": 200,
-      "min_length": 30,
-      "no_repeat_ngram_size": 3,
-      "num_beams": 4,
-      "prefix": "summarize: "
-    },
-    "translation_en_to_de": {
-      "early_stopping": true,
-      "max_length": 300,
-      "num_beams": 4,
-      "prefix": "translate English to German: "
-    },
-    "translation_en_to_fr": {
-      "early_stopping": true,
-      "max_length": 300,
-      "num_beams": 4,
-      "prefix": "translate English to French: "
-    },
-    "translation_en_to_ro": {
-      "early_stopping": true,
-      "max_length": 300,
-      "num_beams": 4,
-      "prefix": "translate English to Romanian: "
-    }
-  },
-  "tie_word_embeddings": true,
+  "tie_word_embeddings": false,
   "transformers_version": "5.0.0",
   "use_cache": false,
   "vocab_size": 32100

generation_config.json

@@ -4,7 +4,7 @@
   "assistant_lookbehind": 10,
   "decoder_start_token_id": 0,
   "diversity_penalty": 0.0,
-  "do_sample": false,
+  "do_sample": true,
   "early_stopping": false,
   "encoder_no_repeat_ngram_size": 0,
   "encoder_repetition_penalty": 1.0,
@@ -15,21 +15,21 @@
   "max_length": 20,
   "max_new_tokens": 64,
   "min_length": 0,
-  "no_repeat_ngram_size": 1,
+  "no_repeat_ngram_size": 2,
   "num_assistant_tokens": 20,
   "num_assistant_tokens_schedule": "constant",
   "num_beam_groups": 1,
-  "num_beams": 4,
-  "num_return_sequences": 1,
+  "num_beams": 8,
+  "num_return_sequences": 5,
   "output_scores": false,
   "pad_token_id": 0,
   "remove_invalid_values": false,
   "repetition_penalty": 1.0,
   "return_dict_in_generate": false,
   "target_lookbehind": 10,
-  "temperature": 1.0,
+  "temperature": 0.8,
   "top_k": 50,
-  "top_p": 1.0,
+  "top_p": 0.95,
   "transformers_version": "5.0.0",
   "typical_p": 1.0,
   "use_cache": true

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bdb556f4a81591a97c2704daf60b4190b357275dbab520a0a1e2d784d7eb106d
-size 891558696
+oid sha256:eab6d229f31e6720b5515aa66638b8f40510040967f41ae5d5fce9f20117dc1e
+size 990214168